Re-Reasoning Ethics: The Rationality Of Deliberation And Judgment In Ethics 0262037696, 9780262037693

Table of contents :
Contents......Page 6
Preface......Page 10
Acknowledgments......Page 16
I Ethics and Reason......Page 18
1 Introduction......Page 20
1.1 Background: Clinical and Philosophical Orthodoxy in Genetic Counseling......Page 22
1.2 The Example: Real Ethical Decision Making after Genetic Counseling......Page 24
1.3 Philosophical Implications......Page 33
1.4 Four Lessons......Page 37
2 The Difficulties of Applied Ethics......Page 40
2.1 Indeterminacy: What Does a Norm Mean?......Page 41
2.2 Conflicts: Which Norm Prevails?......Page 52
2.3 From Substance to Process......Page 58
2.4 The Scope of Morality and the Limits of Formal Bioethics......Page 62
2.5 Conclusion......Page 64
3 The Problems Generalized, Diagnosed, and Reframed in a New Account of Reason......Page 66
3.1 The Parallel Difficulties of Applied Science......Page 67
3.2 Confronting the Dominant Analytic Tradition......Page 74
3.3 Naturalism Frames a Conception of Reason Suitable to the Pursuit of Science, Ethics, and Life in General......Page 82
3.4 Conclusion......Page 89
4.1 Skilled Deliberative Judgment......Page 92
4.2 Four Resources for Rationality......Page 95
4.3 Characterizing Rationality......Page 116
4.4 Conclusion......Page 120
5.1 The Rationality and Morality of Dying Children......Page 122
5.2 Using the Resources of Nonformal Reason in the Vignettes......Page 136
5.3 Moral Effectiveness: Moving beyond Principles......Page 150
5.4 Conclusion......Page 153
II Ethics as Design......Page 154
6 Ethics as Design and Its Two Distinctive Methods......Page 156
6.1 Whitbeck on Ethics as Design......Page 157
6.2 The Deduction-from-Principles Account of Compromise: A Critique......Page 163
6.3 Moral Compromise......Page 166
6.4 Reflective Equilibrium Liberated, Extended, and Enriched......Page 176
6.5 A Shift from Principles to Values......Page 182
6.6 Conclusion......Page 184
7.1 Deliberation as Dramatic Rehearsal......Page 186
7.2 Deliberation as Narrative......Page 189
7.3 Deliberation Shared......Page 196
7.4 Institutionalized Deliberation......Page 205
7.5 Conclusion......Page 208
8.1 Fights, Games, and Debates......Page 210
8.2 Game Theory......Page 218
8.3 Matching Problems and Processes......Page 225
8.4 Conclusion......Page 236
9 Designing Policies......Page 238
9.1 Designing an Ethical Policy......Page 239
9.2 Designing a Process for Policy Making......Page 258
9.3 Conclusion......Page 263
Notes......Page 266
References......Page 294
Index......Page 310

Citation preview

Re-Reasoning Ethics

C

C

Basic Bioethics Arthur Caplan, editor A complete list of the books in the Basic Bioethics series appears at the back of this book.

Re-Reasoning Ethics The Rationality of Deliberation and Judgment in Ethics

Barry Hoffmaster and Cliff Hooker

The MIT Press Cambridge, Massachusetts London, England

© 2018 Massachusetts Institute of Technology All rights reserved. No part of this book may be reproduced in any form by any electronic or mechanical means (including photocopying, recording, or information storage and retrieval) without permission in writing from the publisher. This book was set in Stone Serif by Jen Jackowitz. Printed and bound in the United States of America. Library of Congress Cataloging-in-Publication Data Names: Hoffmaster, C. Barry, author. Title: Re-reasoning ethics : the rationality of deliberation and judgment in ethics / Barry Hoffmaster. Description: Cambridge, MA : MIT Press, 2018. | Series: Basic bioethics | Includes bibliographical references and index. Identifiers: LCCN 2017038168 | ISBN 9780262037693 (hardcover : alk. paper) Subjects: LCSH: Ethics. | Bioethics. | Medical ethics. | Judgment (Ethics) | Reasoning. Classification: LCC BJ1031 .H74 2018 | DDC 170/.42--dc23 LC record available at https://lccn.loc.gov/2017038168 ISBN: 978-0-262-03769-3 10 9 8 7 6 5 4 3 2 1

Contents

Preface ix Acknowledgments xv I  Ethics and Reason 1 1  Introduction 3 1.1  1.2  1.3  1.4 

Background: Clinical and Philosophical Orthodoxy in Genetic Counseling 5 The Example: Real Ethical Decision Making after Genetic Counseling 7 Philosophical Implications 16 Four Lessons 20

2  The Difficulties of Applied Ethics 23 2.1  Indeterminacy: What Does a Norm Mean? 24 2.1.1  The Meaning of “Euthanasia” 25 2.1.2  The Meaning of “Autonomy” 30 2.2  Conflicts: Which Norm Prevails? 35 2.3  From Substance to Process 41 2.4  The Scope of Morality and the Limits of Formal Bioethics 45 2.5  Conclusion 47

3  The Problems Generalized, Diagnosed, and Reframed in a New Account of Reason 49 3.1  The Parallel Difficulties of Applied Science 50 3.2  Confronting the Dominant Analytic Tradition 57 3.2.1  The Critique of Logic as Method 58 3.2.2  Critique of the Western Philosophical Tradition 61 3.3  Naturalism Frames a Conception of Reason Suitable to the Pursuit of Science, Ethics, and Life in General 65 3.4  Conclusion 72

vi Contents

4  The Problems Resolved in a New Account of Reason 75 4.1  Skilled Deliberative Judgment 75 4.2  Four Resources for Rationality 78 4.2.1  Observation 79 4.2.2  Reasoning Methods 86 4.2.3  Creative Construction 90 4.2.4  Systematic Critical Assessment 95 4.2.5  Synergy of the Four Judgment-Improving Resources 97 4.3  Characterizing Rationality 99 4.4  Conclusion 103

5  The Process of Ethical Resolution: Using the Resources of Nonformal Reason 105 5.1  The Rationality and Morality of Dying Children 105 5.1.1  The Rational Discovery of the Diagnosis 106 5.1.2  The Rational Morality of Dying Children 114 5.2  Using the Resources of Nonformal Reason in the Vignettes 119 5.2.1  Ms. B 119 5.2.2  Ms. F 128 5.2.3  K’aila 130 5.2.4  Mrs. Smith 131 5.3  Moral Effectiveness: Moving beyond Principles 133 5.4  Conclusion 136

II  Ethics as Design 137 6  Ethics as Design and Its Two Distinctive Methods 139 6.1  Whitbeck on Ethics as Design 140 6.2  The Deduction-from-Principles Account of Compromise: A Critique 146 6.3  Moral Compromise 149 6.3.1  The Nature of Compromise 151 6.3.2  The Moral Life of Finite, Developing Agents 152 6.3.3  The Rationality of Compromise 158 6.4  Reflective Equilibrium Liberated, Extended, and Enriched 159 6.4.1  Reflective Equilibrium 159 6.4.2  Wide Reflective Equilibrium 160 6.4.3  Liberation, Extension, Enrichment 164 6.5  A Shift from Principles to Values 165 6.6  Conclusion 167

7  Designing Deliberation 169 7.1  Deliberation as Dramatic Rehearsal 169 7.2  Deliberation as Narrative 172

Contents vii

7.3  Deliberation Shared 179 7.3.1  Deliberating with a Doctor 180 7.3.2  Deliberating with a Clinical Ethicist 184 7.4  Institutionalized Deliberation 188 7.5  Conclusion 191

8  Designing Practices, Institutions, and Processes 193 8.1  8.2  8.3  8.4 

Fights, Games, and Debates 193 Game Theory 201 Matching Problems and Processes 208 Conclusion 219

9  Designing Policies 221 9.1  Designing an Ethical Policy 222 9.1.1  The Nature of Tragic Choices 225 9.1.2  Making Tragic Choices 228 9.1.3  Approaches to Allocating Kidneys for Transplantation 231 9.1.4  Assessing the Approaches to Allocating Kidneys for Transplantation 234 9.1.5  Moral Compromise 238 9.2  Designing a Process for Policy Making 241 9.3  Conclusion 246

Notes 249 References 277 Index 293

Preface

This ethics book is unique. Its content is different and distinctive. It is not about the moral theories that are taught in philosophy courses and written about in philosophy journals. (We use “ethics” and “morality” interchangeably.) Moral theories are comprised of principles, and moral reasoning is taken to consist of applying those principles to the facts of problems to deduce right answers. The descriptions of problems are brief, and the resolutions that are logically derived are necessary, certain, and universal. That theoretical depiction of ethics is attractive, but it is too simple, too spare, and too abstract. Telling the truth, for example, can be a challenge. You adhere to principles of telling the truth and preserving life, but you encounter a situation in which lying is needed to save a life. What is the right thing to do? Abide by the principle that prohibits lying or abide by the principle that prohibits killing? Which is the right moral principle, and why is it the right moral principle? More generally, how is the right moral theory to be determined? This book is about real moral problems that are complex, contextual, and dynamic, how these demanding problems are framed, examined, and settled, and how those analyses, constructions, and resolutions can be rational. The conception of ethics presented here is unusual because it is empirical and naturalistic. Examples of moral issues are taken from research in anthropology and sociology. But how can such a depiction of ethics be normative? By recognizing what orthodox moral philosophy ignores: moral judgment. Determining which principles, values, and facts are relevant requires judgment; resolving conflicts between principles and values requires judgment; and applying general concepts to particulars requires judgment. Judgment is ubiquitous in our lives, and ethics is no exception. The normativity of ethics inheres in good judgment, and good judgment emanates from well-designed, rational processes of deliberation. This book is about nonformal reason, a more expansive conception of rationality that

P

P

x Preface

is different from and independent of the formal reason of logic and mathematics and that suffuses our lives. Re-reasoning ethics is about forming rational processes of deliberation that produce rational judgments. That is the best that finite, fallible human beings can do. This book is the culmination of a decades-long collaboration between two philosophers in different fields, one who specializes in bioethics (BH) and the other who specializes in philosophy of science (CH). The common enterprise is not surprising, however, because both the rationality of bioethics and the rationality of science is the nonformal reason that fosters good judgment. The book has two parts. The first part, “Ethics and Reason,” develops the unrecognized nonformal reason that instills a different, augmented rationality into ethics. It is presented by example and by showing how the resources of nonformal reason operate in science, a paradigm of rationality. The second part, “Ethics as Design,” explains how moral compromise and a liberated, extended, and enriched reflective equilibrium are used in designing ethical decision making and how deliberation, practices, institutions, and policies are designed. The contents of the chapters are summarized below. Chapter 1. Introduction. Women who are at risk of having a child with a genetic condition were interviewed about how they deliberated about whether to try to get pregnant. This research provides a real, illuminating example of decision making. The women’s genetic counselors believed that the women should decide by applying the principle of maximizing expected utility to their situations. The women disavowed that approach. Instead, they imagined scenarios of what it might be like for them and their family to live with an affected child, and they assessed their reactions to those scenarios. For moral philosophy, by contrast, applying principles is the rational way to decide. Beginning with this example, this book will explain how the deliberation these women undertake is quintessentially rational. The next two chapters show that the logic-based approach of formal reason creates parallel problems in ethics and in science. Chapter 2. The Difficulties of Applied Ethics. Real cases are used to show that there is much more to ethics than reaching decisions based on deduction from principles and rules. Because the meanings of crucial terms in norms, such as “euthanasia” and “autonomy,” are indeterminate, they have to be clarified and specified. When principles and rules conflict, how is the one that prevails to be determined? What makes a problem a moral problem, and what does being a moral problem mean? After nonformal reason is introduced in chapter 4, chapter 5 discusses how it can manage several of the cases presented in chapter 2.

Preface xi

Chapter 3. The Problems Generalized, Diagnosed, and Reframed in a New Account of Reason. The problematic features of moral situations revealed in chapter 2 have exact counterparts in scientific research. Given that, it is not unreasonable to expect them in every domain. Their appearance in science makes it especially clear that the problem is located in the common conception of reason as logic that currently underlies both ethics and science. To prepare the way for examining the nature of an appropriate rational procedure, chapter 3 surveys the received Western tradition of characterizing rationality formally; then it explains why this conception of rationality cannot handle serious problems and why a radically different approach is required. The natural framework for addressing these matters is a philosophical naturalism that recognizes that humans are fundamentally finite and fallible but possess a powerful ability to learn. This start refocuses rationality around intelligent problem solving. The following two chapters develop a nonformal conception of rationality, first for science (where the issues are more cleanly drawn) and then more broadly. Chapter 4. The Problems Resolved in a New Account of Reason. Deliberative judgment formation, or simply judgment, is a core human skill that is largely irreplaceable by rational formalisms. Indeed, it is required to create those formalisms. Judgment is rationally learnable and improvable like other skills; and, like all intelligent creatures, humans need abundant practical experience with contextual problem solving for judgment formation to be sufficiently relevant, attentive, perceptive, and judicious to support intelligent problem solving. Equally telling for the role of judgment, even in learning about our norms, is that norms and normative methods have been massively elaborated historically in both science and ethics as experience has been gained, as is characteristic of knowledge and skills that have to be acquired by the human species through learning. Rationality can itself be learned and improved through experience in problem solving, for example, in the way an experienced lawyer powerfully marshals precedents and analogies. The same self-improvement characterizes an increasingly powerful scientific method, the spread of institutional ethical resolution processes, and so on. In exercising reason, there are four principal resources or means for the improvement of skilled judgment: observation, the use of both formal and informal reasoning procedures, constrained but creative construction, and systematic critical appraisal. These four bundles of processes are utilized both by individuals and, typically more powerfully, by communal groupings such as research teams. The bulk of chapter 4 is devoted to expositions

xii Preface

of the four resources of nonformal reason and their strategic deployment in problem solving. The discussion closes with responses to these questions: (Q1) What kind of account is to be given of the rationality of a problem solution or a decision? Response: that it was formed through a properly designed deliberative process. (Q2) What is the content of an account of rationality? Response: a properly designed deliberative process utilizes each of the four judgmentimproving resources, along with the tools of compromise and reflective equilibrium formation, as feasibly as possible. And (Q3) What should be the relationship between an account of rationality and an account of our empirical cognitive capacities? Response: naturalists hold that both the normative and the empirical accounts are fallible and learnable and as such must be mutually adapted in a respectful two-way iterative adjustment. Adaptation of each is driven by identification and avoidance of failure and adoption of success: the empirical account by success in understanding the actual problem-solving processes humans use, both its strengths and its weaknesses, biases and limitations; and the normative account by success in guiding the investigation of problem solving, improving our investigatory capacities in the process. The next three chapters consolidate this account of nonformal reason by showing how it deals with the cases presented in chapters 1 and 2, adding two distinctive tools for constructing problem resolutions—compromise and (enriched) reflective equilibrium—and illustrating rational deliberation in more examples. Chapter 5. The Process of Ethical Resolution: Using the Resources of Nonformal Reason. Real examples illustrate how the resources of nonformal reason are used practically. The first is new. It comes from Myra Bluebond-Langner’s ethnographic study of children with leukemia. These children desperately wanted to know what was wrong with them, but their parents and the health care staff remained silent. The children used the four resources of nonformal reason to discover that they were dying. And they remained silent. Understanding why presents us with a poignant, magnificent example of intelligence. The resources of nonformal reason then are used to illuminate the plights of Ms. B, Ms. F, K’aila’s parents, and Mrs. Smith described in chapter 2. Chapter 6. Ethics as Design and Its Two Distinctive Methods. The development of nonformal reason is completed with Caroline Whitbeck’s application of design in engineering to the design of moral problem solving, supported by the incorporation of two distinctive methods for rationally designing resolutions of practical problems. From Whitbeck’s work the

Preface xiii

notion of ethics as design for flourishing is developed, thereby expanding the scope of ethics to encompass ethical problem solving from the personal to the political. One of the distinctive methods for solving ethical problems constructs a fully engaged moral compromise, which requires that the unmet demands of the conflicting ethical claims that are compromised are acknowledged, and thus retains the full richness of ethical considerations for future deliberation. The other method creates a wide reflective equilibrium that is liberated (rather than confined to inference and formal theory construction), extended (to encompass diverse forms of moral deliberation), and enriched (to encompass creative moral options of all kinds). Chapter 7. Designing Deliberation. Deliberation occurs whenever intentional judgment formation is required but especially when important decisions rife with uncertainty, complexity, and incommensurability have to be made. The best way to understand deliberation is by examining the operation of nonformal reason in various examples. The examples of chapters 1 and 2 provide instructive exemplars, and more are presented in chapter 7, drawn from social science research. Deliberation can take the form of dramatic rehearsal or a narrative, for example, storytelling by occupational therapists trying to understand the behavior of a young patient. A doctor and a patient can engage in shared deliberation, and doctors and nurses in a neonatal intensive care unit can use institutional deliberation to discuss their disagreements. When participation is properly designed, deliberation can substantially reduce alienation by helping to reunite reason and motivation. Doing that is inherently ethical, and it also is the foundation of the professional ethics of relationships with patients/clients and other professionals. The final two chapters expand the scope of ethical design for flourishing from the conduct of professionals to the formulation of public policies. Chapter 8. Designing Practices, Institutions, and Processes. Ethics is embedded in the practices and institutions of society. Three examples illustrate the communal importance of ethical design. One is Anatol Rapoport’s classic illumination of fights, games, and debates. Another is an examination of the Prisoner’s Dilemma in game theory. And the last is Lon Fuller’s account of institutional design, invoking the notion of “polycentric” problems and different forms of governance, especially mediation, in designing solutions. Chapter 9. Designing Policies. Designing ethical policies is illustrated with two real cases. One is motivated by Guido Calabresi and Philip Bobbitt’s book on Tragic Choices. A tragic choice is forced when fundamental values conflict and none of the values can be abandoned. Because cadaver kidneys for transplantation are scarce, decisions about how the kidneys will be allocated have to be made. Neither of the two conflicting values,

xiv Preface

equality and efficiency, can be renounced. The resolution requires cycling, a process driven by the remaining residue of which value currently is being repudiated more extensively. Cycling shifts the priority of the two values back and forth over time. This example also exemplifies the necessity and the importance of moral compromise, which is anathema to formal moral theory. The other example is a brief presentation of Anthony Culyer’s illuminating account of how the National Institute for Health and Clinical Excellence (NICE) in England and Wales designed a deliberative process for assessing new health care technologies.

Acknowledgments

Excerpts and/or adaptations from five articles that we have published in journals appear in this book. In chapter 1 an example from a publication in Bioethics (Hoffmaster and Hooker 2009) that shows how women decide whether to try to get pregnant after receiving genetic counseling is used to appreciate what real ethical decision making is. In chapters 3 and 4 an extended publication in Axiomathes (Hooker 2010) is used to develop a critique of logic-based rationality (chapter 3) and an alternative account of nonformal reason (chapter 4). Chapter 5 presents an article from the Hastings Center Report (Hoffmaster 2011) that depicts searches conducted by children with leukemia who want to know what is wrong with them to illustrate the rationality of these non-formal procedures. In chapter 6 an article in Social Theory and Practice (Hoffmaster and Hooker 2017) portrays a disagreement between a nurse and a doctor in an intensive care unit to show how a compromise can be constructed, supplemented by an explanation of how moral compromise can be rational. Chapter 9 is about designing policies, and the primary example is the development of a policy for allocating the scarce supply of cadaver kidneys for transplantation, published in an article in the Milbank Quarterly1 (Hoffmaster and Hooker 2013). We are pleased to have had our work appear in these journals, and we are grateful to have this work made available here. We have been fortunate to be guided by Philip Laughlin, our editor at the MIT Press. He received a novel and creative, but expansive and discursive manuscript, and he spurred the authors to turn it into a tighter, more integrated book. We thank him for his help and his attention. We also thank the three anonymous referees who wrote acute criticisms, suggested additional issues to examine, and recommended extra readings. We certainly need to thank our families too, especially given how long we have been working on this book. Barry is grateful to Sylvie, his wife, for her understanding of what this project has meant to him, and, even

A

A

xvi Acknowledgments

more, for her acceptance of the long time it has taken. He hopes his children, Samantha and Nicolas, might find some of their father in this book. Cliff thanks wife Jean for her unfailing patience, interest, and support and grandchildren Selkie, Ursula, and Arthur for keeping him young at heart and with a well-stocked immune system.

I  Ethics and Reason

E

E

1  Introduction

C

What is “re-reasoning ethics”? Re-reasoning ethics is about developing and defending a conception of rationality for ethics that is expansive, flexible, and effectual. It is independent of deductive reasoning, a paradigm of rationality that dominates principle-based moral philosophy. The conclusion of a deduction is necessary, certain, and universal, and those are the features that propagate the universalizability that orthodox moral philosophy demands of ethics. In moral philosophy ethics cannot be contingent, subjective, or relative. If it is wrong for a teacher to criticize a student in front of the class, it is wrong for every teacher. When deduction is taken to be the rationality of ethics, morality becomes a moral theory comprised of moral principles that serve as the major premises of deductive reasoning. A moral theory can have one supreme foundational principle, for example, Kant’s categorical imperative or act utilitarianism, or a plurality of principles, for example, intuitionism. Assimilating ethics to deductive reasoning, principles, and moral theories produces an attenuated rationality and a constricted ethics. This influential and widespread, but misguided, depiction of ethics is the result of condensing rationality to deduction. Problems quickly proliferate in an ethics of applying moral principles. How is a principle justified? By deducing it from a higher principle? If so, how is the regress stopped? How is the relevance of a principle determined? How are the general concepts in principles to be construed? How is the relevance of facts ascertained? How are conflicts between principles resolved? Only after all of these questions are answered could deduction give answers to moral problems by subsuming facts under principles. Moral philosophy has no answers to these questions. Moral problems have to be identified, framed, analyzed, and assessed before a judgment can be arrived at about what to do. That requires deliberation. Potential ways of resolving problems have to be surveyed and evaluated. Relevant issues are interrelated, so the ways they impinge on

I

4 

Chapter 1

one another have to be identified and assessed, and the depictions of the issues have to be constantly examined and adjusted. All the elements of the deliberation and their interactions have to be scrutinized. Is the problem framed properly? What responses could handle the problem? What are the strengths and the weaknesses of the various resolutions? How should the options be compared? How can a decision be made? When should the deliberation stop? These matters, and endless others, require judgment. But how can judgments be rational? They are not deduced. The rationality of judgments inheres in a more expansive conception of rationality. How can there be rationality beyond deduction? The rationality of deduction resides in the form, not the content, of an argument. For that reason we will call this kind of rationality formal reason. If rationality were all and only formal reason, then nothing but logic and mathematics would be rational. There would be little rationality in our lives. Yet we make judgments ceaselessly, often unknowingly. Many are trivial, some are important, a few are momentous. Sometimes we make judgments carelessly or frivolously, but at other times we make judgments carefully and rationally. When we have to make hard decisions, we examine the options extensively, we deliberate seriously, and we make judgments. We cannot be sure that we made the right decision, but we know that we made a rational decision. A judgment is rational when it emanates from a rational process of deliberation. We will call this kind of rationality nonformal reason. In chapters 3 and 4, the core of this book, nonformal reason is explained and defended by showing that the rationality of science, a field that is regarded as eminently rational, is nonformal reason. Because the orientation of this book is empirical and naturalistic, many real examples are used, not just to elucidate but to vindicate. A real example is presented in this introductory chapter. It comes from research about how women who are at risk of having a child with a genetic condition decide whether to try to get pregnant. The women’s genetic counselors believe that their clients should decide by applying the principle of maximize expected utility to their situations, and that doing so would give them the right answer. The women discard that principle. They design a process for deliberating about what is meaningful to them and what they should do. This book explains how their deliberation is rational and how, when deliberation is rational, a judgment that emanates from that deliberation is rational. To be more explicit, we use “reason” to refer to a conception of the nature and scope of rationality, and use “reasoning” to refer to the specific activities undergone in exercising rationality. We use “formal” and “nonformal”

Introduction 5

reason to name two different conceptions of rationality, and “formal” and “informal” reasoning to refer to two distinct kinds of activities that may occur under them. In this scheme formal reason contains only formal reasoning, while nonformal reason includes both formal and informal reasoning. We take the reasonings focused on or organized around a particular subject or problem to be deliberations. Deliberations are rational when they are appropriately organized (see chapter 4). All judgments issue from deliberations, however brief or extended, explicit or tacit. Judgments are rational insofar as the deliberations that gave rise to them are rational. Judgment can be improved by learning how to improve deliberation; this includes judgments about improving rational deliberation (see chapter 4). 1.1  Background: Clinical and Philosophical Orthodoxy in Genetic Counseling The following vignette recounts a meeting among three genetic counselors, Bill Smith, Joseph Giordano, and Albert Samuels, at which Charles Bosk, a medical sociologist studying their work, also is present (Bosk 1992). It is a Friday afternoon at postclinic conference. Bill Smith is providing some follow-up on the Meullers, a couple who had recently been in for counseling. Bill explains that the Meullers are a couple who have recently had a child with Von Recklinghausen’s disease. Mr. Meuller has four café au lait spots (a marker of this dominant condition in its less serious manifestations). He has been seen by a number of physicians, two of whom thought the spots meant that he carried the gene. Bill then asked Giordano to continue. He first described the counseling session he had had with them and then described a conversation with the Meullers’ pediatrician. The pediatrician knew how much Mrs. Meuller did not want an abortion and was angry at the clinic for assigning carrier status to Mr. Meuller without a skin biopsy. Giordano continued: “To make a long story short, Mr. and Mrs. Meuller went shopping for a dermatologist who said that the spots did not definitely mean that he carried the gene. Mrs. Meuller called today to say that she was going to keep the pregnancy.” Giordano, Smith, and Samuels are all distressed that Mrs. Meuller is going to carry this pregnancy to term. They all believe that there is at least a 50 percent chance that the infant will have the disease (although the seriousness of expression ranges from Mr. Meuller’s café au lait spots to the features of the Elephant Man.) Giordano sums up with an expression of helplessness. Having sized up his options, he says, “The only thing I can accomplish if I intervene is harm.” Bill Smith says, “I want to sum up for your benefit, Bosk. This is a couple who have chosen to minimize and deny the risk, who are bound and determined to

6 

Chapter 1

keep the pregnancy. This is a different situation than the Allegras’ (other clinic patients), who are trying to decide whether or not to take the risk. These people (the Meullers) are ignoring it entirely. There are three different situations to worry about. Cases where parents accept a very high risk, cases where they deny there is any risk at all, and cases where they can’t get the meaning of the risk clear in their own minds.” (Bosk 1992: 46–47)

This vignette displays three central features of the way genetic counseling is perceived and conducted. First, genetic counseling is about risk. But what does “risk” mean here? How are these genetic counselors using the term when they talk about “taking the risk,” and the like? It is useful to distinguish two notions of risk. In a narrow sense, risk is simply the probability that a child will be born with a particular genetic disease. In a broad sense, risk is about not just the probability of that bare event, but also includes the manifold ramifications of what it could be like to live with a child who has that genetic disease. Some doctors may consider that only narrow risk is relevant, but it is clearly wide risk that ultimately matters to patients. Second, uncertainty pervades the determination of both risks. In this vignette the probability of a child having a genetic disease is not clear because it is not known whether Mr. Meuller carries the gene for Von Recklinghausen’s disease. That uncertainty is compounded by the widely varying and unpredictable expression of Von Recklinghausen’s disease. In its mild form the disease causes areas of hyperpigmentation (café au lait spots) on the skin. Individuals without the disease also can have this hyperpigmentation, usually in milder form. In its extreme form the disease can cause grotesque deformities,1 mental retardation, epilepsy, skeletal abnormalities, gigantism of a limb, and a variety of tumors. Moreover, for reasons that remain unknown, the disease tends to be more severe when transmitted by a male. So even if Mr. Meuller were a carrier of the disease and if his child were to be born with the disease, how severe the disease would be in that child is unclear. Third, people who receive genetic counseling are expected to make their subsequent decision on the basis of the (wide) risk they face. The assumption is that when the risk is too high, they (a couple or a woman) should not try to conceive or, if they already have conceived, they should not continue the pregnancy. Giordano, Smith, and Samuels are upset because they believe that the risk for the Meullers is too high and consequently that the Meullers’ decision to continue the pregnancy is not justifiable. Here “risk” is understood in the terms of standard decision theory: First identify a range of narrow risks (probabilities) that describe the relevant

Introduction 7

range of outcome possibilities and describe these outcomes, then (assuming value is unaffected by probabilities) determine the value of each outcome to the affected party, and multiply it by its narrow risk (probability) to yield (narrow) risk-weighted (or “expected”) outcome values (utilities) for each option. The standard decision principle is, then, to choose the option with the highest or maximum expected utility (MEU). In the present case, whether to create or retain a pregnancy should be decided by choosing the outcome that would produce the greater of (a) the expected value (utility) of leading a childless life or of adopting a child (which will have its own wide risks) and (b) the expected value of having a child, but one having the genetic outcome for that option. By tacit agreement this also becomes the morally appropriate decision. The effectiveness of genetic counseling is correlatively assessed in terms of how closely the ensuing decision comports with the result of the MEU process (assuming “typical,” or perhaps clinically assessed, utilities for the Meullers). Smith, Giordano, and Samuels evidently are troubled by the decision of the Meullers because they cannot understand how the Meullers’ decision maximizes their expected utility. They conclude that the Meullers have failed to grasp and use the proper decision-making procedure. But, strictly, this is beyond their competence. The standard formulation of the MEU principle incorporates a sharp distinction between facts (the descriptions of wide risks) and values (their evaluation), assigning them separate and independent roles. That tacit theoretical representation of the dichotomy in turn structures the practice of genetic counseling and defines the roles of its practitioners. It is the source of the widely espoused view that genetic counseling must be “nondirective,” which limits the responsibility of a genetic counselor to providing only facts, i.e., probabilities and physical consequences. Determining the relevant values and the importance of those values is the sole responsibility of those being counseled, so genetic counselors should in no way try to influence or direct them. How well does the MEU approach to decision making suit the plights of those who receive genetic counseling? How do they evaluate the wide risks—i.e., probabilities (narrow risks) and their outcomes—explained to them? How, as Smith puts it, do they get the meaning of risks “clear in their own mind”? How do they deliberate and make their difficult decisions? 1.2  The Example: Real Ethical Decision Making after Genetic Counseling The work of Abby Lippman and F. Clarke Fraser reveals how radically the actual decision-making processes of women at risk of bearing a child with a

8 

Chapter 1

genetic condition depart from the prescriptive MEU model.2 We will focus on the women because they were the principals in this research. How do these women conceive their problem and work their way through it?3 They begin by minimizing or ignoring the probabilities conveyed to them by genetic counselors, instead formulating their situation in binary terms: “Either we will have an abnormal child or we will not” (LippmanHand and Fraser 1979b: 57, 1979d: 332). Even if the denominator in the probability they are given is large, the women focus on the 1 in the numerator and realize that their child could be that one. Two women explain: Even if it’s a 6% chance … there will still be that chance of a kid eventually having it. I know you doctors have a very good heart, you want to convince us there is always a better chance if you look after this and that. But there is always a chance that it can flip and [will happen]. This scares me. (Lippman-Hand and Fraser 1979a: 120)

The critical factor for these women is a residual, ineliminable possibility of having a child with a genetic condition. Whether the probability is relatively high or relatively low, that prospect cannot be removed—the 1 in the numerator always will be there—and that is what is relevant to them. This feature persists however much the uncertainties are reduced: even if it were possible, knowing the precise probabilities and correlative consequences would not change the necessity of facing up to the possibility of that outcome.4 How, then, do the women respond to this invariant feature of their predicament? They construct what Lippman and Fraser call “scenarios”— imagined accounts of what it would be like to have a child with the presaged genetic condition—and assess whether they could cope with the worst scenarios they can envisage. Two factors figure prominently in the way women apprehend their problem and thus in how they formulate and react to their personal scenarios: an appreciation of the amorphousness and complexity of their plight, and a sensitivity to the burden of having to make a decision. The situations in which these women find themselves are ill-defined, fuzzy, and incredibly complicated. They do not know how serious their child’s condition, should it eventuate, would be because of the variable expressivity of genetic diseases. They must consider everyone who is likely to be affected—the child, themselves, their partner, extended family, friends, and society in general—surmising what the plausible, but often uncertain, effects on these people could be—including transformation of

Introduction 9

their self-images, goals, and values—and what range of responses might be acceptable and what would not, and then appraise those effects. From their own perspectives, they must work through how the multifarious outcomes could impinge upon them in their roles as parent (or potential parent), spouse, relative, and member of various social networks. The sheer number, variety, and complexity of the factors they need to consider are compounded by the shifting interactions among them and the incommensurability of their evaluations. Moreover, there are no established social norms to structure or guide their decision making. Because the technologies that create these problems are comparatively recent and rapidly developing, because this kind of decision is relatively uncommon, and because public discussion of the issues surrounding such decisions is muted, social guidelines have not evolved and crystallized. Genetic counselors commonly are asked for guidance, along the lines taken by one couple: This is a high risk, one in four. Are there others like us? What do people do with this risk? Do they take the chance? What do doctors suggest they do? We’ve never had to face this kind of thing before, so [we] want to know what others say. (Lippman-Hand and Fraser 1979a: 121)

Such a request can be regarded as a flight from responsibility, as an attempt to relinquish autonomy and avoid the burden of having to decide, and, given how onerous the decision is and how much is at stake, it indeed might be that for some women. Often, however, there is a more charitable interpretation. Those who make this kind of request are not seeking to be told what to do or soliciting approval of their decision. Rather, they are searching for missing information that can help them frame their problem and give them a sense of what courses of action are reasonable. They are seeking societal direction or standards that can help them reduce their uncertainties and understand their options: Once they [the women] have expressed the uncertainties and ambiguities confronting them, their requests for advice may be seen as a search for cues to the range of behavior that might be appropriate in their novel circumstances. In the many spheres of life in which one participates, there are rules and constraints imposed on one’s activity by outside forces as well as a variety of role models against which one can judge “what is done” in specific situations. These cues orient the actor in choosing a course of behavior. Parents facing the chance of having an abnormal child do not have such guidelines for their reproductive behavior. Awareness that all who undertake childbearing risk having an abnor-

10 

Chapter 1

mal offspring is at best subliminal, and parents are unlikely to know of others in a situation similar to theirs whose behavior might provide some sort of social framework within which they can consider their own options. Thus, these questions may be a search for such guidelines. Even when the question is put more directly—“What would you do?”—it may be part of this search for what would be socially acceptable and reflect the desire to explore the limits of their alternatives. (Lippman-Hand and Fraser 1979a: 124)

What the women desire, in other words, is help with framing and structuring an amorphous, complicated, messy problem and with delimiting the range of acceptable responses. Understood in this way, their requests are an attempt to exercise responsibility, not abdicate responsibility. Yet as these women create and explore their scenarios, they discover a plurality of moral norms relevant to their plight. The principal norms pertain to not causing pain and suffering and to the multiple responsibilities of parenthood. The latter include the responsibilities of becoming a parent and being a parent to a child, the responsibilities of parents to other members of their families, the responsibilities of parents to society, and the responsibilities of parents to themselves. There are, however, no higher norms that identify which norms, of all the ones a woman might survey, are strongly relevant to this particular problem; that determine the comparative importance of the norms that are relevant; and that resolve conflicts between relevant norms. Nor are there higher norms that direct how relevant norms are to be elaborated and specified as the details of scenarios multiply and mutate or how challenges to the correctness of relevant norms are to be assessed. In addition, there are no methodological norms that govern how to conduct inquiries and deliberations about these matters rationally. At the same time, the women are acutely aware of the burden of being forced to make this kind of decision. Although there is a general background risk of abnormality for all children who are conceived, women who receive genetic counseling face a heightened risk that renders their decision making more difficult and that exposes them to special scrutiny from their family and friends, from members of the health care professions, and from society. The onus of having to make a choice that most other women do not have to make is itself a significant component of their difficulty. Thus these women not only confront a momentous and seemingly intractable decision, they also have to assume the explicit responsibility that attaches to whatever decision they make. How do these women work through the scenarios they imagine? Strategies of simplification and neutralization help them to defuse the

Introduction 11

amorphousness and complexity of their situation, to construct their problem in a manner that makes it manageable. This involves, as we have seen, reducing it to binary terms and trying out worst-case scenarios to see whether they are acceptable. Because these women are highly motivated to have a child, or another child, their problem becomes one of whether they can neutralize the worst outcomes they can imagine. They are searching for a reproductive “green light,” and to get that green light, they must feel that they are able to cope with their worst-case scenarios and able to accept responsibility for such a risky undertaking. Two factors help these women come to grips with their uncertainties. One is the presence of an unaffected child. A woman who already has an unaffected child realizes that she need not “lose,” i.e., lose the experience of raising an unaffected child, if she takes a chance, and that reassurance makes it easier to go ahead. The existence of an unaffected child also relieves the pressure on women to “normalize” their biological and social roles as mothers. Finally, insofar as a mother considers that she is acting in the interests of another child who needs a sibling, an unaffected child can help diffuse the feeling of responsibility for a decision to proceed. The other factor is what a woman already has gone through, either as affected by the genetic condition herself or as the mother of an affected child. These experiences concentrate and constrain the imaginative construction of scenarios and provide a touchstone for assessing one’s ability to cope with the possibilities that are canvassed. Living with a condition enables a woman to devise more detailed, focused scenarios and to conclude either, “I’ve been through it and it’s not so bad,” or “Nobody should have to go through what I’ve been through.” The birth of an affected child, regardless of whether that child lives or dies, strengthens the disposition to see outcomes in binary terms but also provides a basis for estimating whether one could endure a repetition of the experience. Thus these two kinds of experience can annul some of the pervasive uncertainties that confound these women and make it easier for them to decide one way or the other. But only some women have those experiences. Women who feel they could cope with the most dire problems they can imagine of caring for an affected child, as well as with the responsibility of bringing about the birth of an affected child, try to conceive. Women who feel they could not manage one or both try not to conceive.5 The actual decision making by these women is a dramatic departure from applying the MEU principle, but it is not a surprising departure because the prescriptive MEU model does not serve them well. The MEU principle delivers clear decisions only when both the suite of possible outcomes and the

12 

Chapter 1

probability for each of those outcomes are known and the value of each outcome is clear; in more realistic situations one or more of these elements is often unavailable. For example, genetic counselors can often provide only vague or conditional rates (probabilities) and outcomes, either because it is impossible to make a firm diagnosis, because a diagnosis cannot be associated with a single hereditary pattern, or because a person’s genetic status cannot be definitively determined. In those circumstances, the MEU model provides little or no purchase on the problem. More important, even if precise numerical rates and outcomes could be provided, the real issues still would not be addressed. As we have learned, the women generally ignore the probabilities or convert them into yes/no options. They restate and interpret the percentages they have been given to emphasize the possibility that cannot be dispelled—having an affected child—and it is this enduring possibility that is most salient to their decision making. Nor should the making of clear value assessments be treated as unproblematic, for example, by considering it a private matter for each woman. The decisions these women have to make engage a far wider range of values than just those that concern the condition of the child. Such a broad span of values typically is not represented in the simplistic choice models offered in genetic counseling. From this perspective as well, the MEU decision framework used in counseling proves deeply impoverished and consequently provides the wrong orientation to this decision making. Nor will expanding the scope of the MEU principle alleviate these problems, for two important reasons. First, genetic counselors are not equipped to provide probabilities for how partners and others involved will respond, or for how women will come to feel about themselves after giving birth, or for an array of other broader consequences. Neither, in general, is anyone else better placed to do so. Second, and more profoundly, many of the values involved are ones that the women themselves have never considered, let alone consciously examined and accepted. Now they have to construct and appraise these values for themselves as their inquiry proceeds. They are forced to consider why they want to have a child. Why is it important to the woman to be a mother, and how genuine is that goal? How much would she be doing this for herself, and how much to please or satisfy others or to fulfill societal expectations? Each woman has to assess other matters that are important to her. How deep and durable is a commitment to a spouse or partner, and how might caring for a child with incessant emotionally and physically draining needs jeopardize that relationship? How much does a job or career mean to her? Would she lose raises in salary and miss promotions because of unremitting demands on her time? How strong

Introduction 13

are her bonds with her family, and how much help could she reasonably expect from relatives without alienating them? How much does she prize independence and control, and how extensively might she become reliant and dependent on others, including busy health care professionals and strangers? Each is forced to consider the expectations she has for a child. What kind of person would she like her child to be? What achievements would make her proud? How would she react to a child who, in many respects, might not be like most other children? How could she cope with the teasing and taunting that might be inflicted on someone who is particularly vulnerable yet completely innocent and, most important of all, is her child? These women have to scrutinize, honestly and brutally, whether they have the resources to be the mother of a child who has a genetic condition. Do they have sufficient patience, courage, strength, and humility? Can they make the sacrifices that would be required, and what, in the end, would make those sacrifices worthwhile? It is hard to even identify all the discrete values these questions implicate, let alone claim that those values already exist in a determinate form and are quantifiable in a fashion that would allow them to be multiplied with numerical probabilities. Decisions about motherhood produce the usual hopes and dreams alloyed with the usual worries and fears. Above all, though, these decisions are about love, commitment, and altruism. These matters might not have been contemplated before, certainly not combined as they now must be and in as much depth as they now must be. The prospect of having a child with a genetic condition precipitates a fundamental reappraisal of what a woman’s life has been about and what she has regarded as important. To propose that these women should be applying their utilities to probabilities distorts and trivializes what is at stake. They are not plugging values into a formula. These women are deciding what kind of person they want to become, what kind of social matrix they want to live in, and whether they and their families have the endowments and the resources to accomplish that. Rather than applying clear and determinate antecedently existing values, they are creating, modifying, and exploring values. It could be contended that the MEU principle is to be applied only after these value identification and assessment processes are completed, but doing that would be needless because by then it would have become clear whether there is a “green light” to proceed, and whatever finer discriminations MEU might provide would, at best, be superfluous. At worst, trying to apply the MEU principle would prove a diversion from the crucial work of sorting out value issues, and the results could not be trusted at any place where they might suggest a different decision.

14 

Chapter 1

The counselors themselves recognize that, even when the numbers are definitive, for example, a 10% chance or a one-in-six risk, their implications are vague. The numbers must be imbued with meaning, and the women do that by incorporating them into overall depictions of their problems: In the absence of certain knowledge of whether or not a future child would be affected, the important issue became what having an affected child would mean. Moreover, one’s numerical risk only had operational meaning when it was perceived in this context. It was not so much that the parents’ perceptions of a diagnosis as severe or otherwise influenced their interpretation of a rate figure, but that this probability was, in itself, seen as just one of the consequences of being “at-risk” (in a general sense) for having an abnormal child: Along with the severity of a condition and its duration, its chance of happening was one of its characteristics. And in incorporating rate information this way parents again seem to be attempting to make it personally meaningful. What could happen, with all its implications and uncertainties, was the issue to be faced. (LippmanHand and Fraser 1979a: 121)

As Lippman and Fraser point out, both counselors and those who have to decide recognize that numerical rates have to be interpreted and that “putting a rate ‘into perspective’ involves closely integrating it with the consequences it entails and the life circumstances of the family involved— something far beyond the simple provision of facts, and a reflection of the need for this kind of information to be processed to become meaningful” (Lippman-Hand and Fraser 1979a: 125). Lippman and Fraser rightly conclude that this kind of understanding requires a different account of decision making: The constant interaction of risk and consequences … suggests that this is an irreducible complex to be processed in subsequent reproductive decision-making and supports the need for a model of such decision-making that involves the heuristic processing of complex information rather than one that assumes a traditional cost-benefit (risk-burden) approach. The latter may be more amenable to formal decision analysis; the former seems more characteristic of what really happens. (Lippman-Hand and Fraser 1979a: 124)

The MEU principle cannot capture the complex, dynamic, contextual, and holistic nature of the problem these women confront because it reduces the formulation and resolution of the problem to two discrete factors that are defined in simple terms and operate independently. Forced application of the spare, albeit elegant, risk-neutral MEU formula does not just crimp and distort what is at issue, it robs the problem of its deep moral import and its personal momentousness.

Introduction 15

This account of how women actually make reproductive decisions after genetic counseling reveals the manifold ways in which their interpretation of information, their framing of problems, and their decision making are contextual. Above all, they are struggling to make the information they receive meaningful by personalizing it to themselves and their situations. To accomplish that, they must put the information in a historical and social context: they must integrate it into their own life histories and the life histories of those close to them. The significance the information assumes will depend, for example, on whether a woman (or someone close to her) is affected by a genetic condition. If so, that experience can give shape and substance to the scenarios that are imagined, and how burdensome that disease has been can influence reactions to those scenarios. The information also must be put in a family context. Whereas a woman who has no children might feel pressure to “normalize” her role as a mother, a woman who already has an unaffected child would not be similarly influenced. On the other hand, a woman who already has a child might feel pressure to provide a sibling with whom that child can play. And the information must be put in a societal context. What resources does society provide to help those who care for a child with a genetic disease? How does society expect these decisions to be made? How do people in similar circumstances decide, and what social norms exist? These women confront one of the hardest decisions a person can make. Their problem is difficult not just because of all that is at stake, but because it is radically open and indeterminate. Their task is not simply to solve a problem but first to formulate it: to structure their problem and then to devise strategies for approaching and settling it. Moreover, there is a dynamic, reciprocal relationship between attempted formulations of their problem and attempted resolutions of their problem. An effort to crystallize the problem will suggest ways of resolving it, but pursuing those resolutions will ramify back into the reformulation of the problem, and that in turn may require revising the approach. For instance, strategies for dealing with a particular scenario (say, a father not coping) will generate new considerations (what kind of help would the father be?) and insights (extended family and other social resources may be important to outcomes) that lead to modifications and extensions of the scenario and to alternative scenarios (e.g., divorce), and features of these in turn will often suggest new aspects of the problem not yet properly captured and new methods for doing so. Formulation, method, and resolution are interdependent and reciprocal processes. In addition, these women have to find a way of stopping this dynamic exchange. They have to figure out what kind(s) of outcome(s)

16 

Chapter 1

would be satisfactory and how to recognize them. All this forces the women to construct new conceptions of themselves as persons and to try out those conceptions. They must determine whether a particular conception of what would give their life worth and integrity is possible for them—whether that conception fits their idiosyncratic histories and their manifold potentialities—and whether a particular conception is desirable for them—whether it demands sacrifices that are too onerous and whether it allows them to realize some of their hopes and dreams or at least some of their aspirations. 1.3  Philosophical Implications Women who receive genetic counseling struggle with what Plato and Aristotle recognized as the fundamental moral question: how ought one to live? Now, however, the subject matter of ethics is different. Edmund Pincoffs regards the prevailing conception of ethics as “so general that it would be tedious to document it”: It is that the business of ethics is with “problems,” that is, situations in which it is difficult to know what one should do; that the ultimate beneficiary of ethical analysis is the person who, in one of these situations, seeks rational ground for the decision he must make; that ethics is therefore primarily concerned to find such grounds, often conceived of as moral rules and the principles from which they can be derived; and that meta-ethics consists in the analysis of the terms, claims, and arguments that come into play in moral disputation, deliberation, and justification in problematic contexts. (1986: 14)

Contemporary philosophical ethics is preoccupied with justifying correct answers to moral problems, and correct answers are those that can be derived from moral principles or rules. This method provides what Pincoffs calls the “rational ground” for a moral decision. We recognize it as the formal logic-based or analytic approach to ethics that we introduced at the outset. “Quandary ethics,” as Pincoffs dubs this logic-based, formal-reason brand of morality, is alien to much of moral philosophy: “Plato, Aristotle, the Epicureans, the Stoics, Augustine, Aquinas, Shaftesbury, Hume, and Hegel do not conceive of ethics as the quandarists do” (1986: 14). For Aristotle, for example, “In ethics the leading question concerns the best kind of individual life, and the qualities of character exhibited by the man who leads it.… Moral problems are given their due but are by no means stagecentre. The question is not so much how we should resolve perplexities as how we should live” (Pincoffs 1986: 15).

Introduction 17

The women whom Lippman and Fraser studied certainly are trying to solve a problem, but they are not approaching it formally in the manner of a Pincoffs quandary. Rather, their age-old problem is precisely how they should live—they are deciding what they need to do to flourish. What is the best life for them, and for their child should they have one, and for their families? What qualities of character does that kind of life require? Are their families and they capable of exhibiting those qualities? To be sure, there is a particular problem they must resolve, but that problem is embedded in and subsidiary to their more sweeping moral inquiry. As these women astutely realized, their problem is satisfactorily resolvable only within the broad contexts of their inquiry, and their inquiry engages their entire being, not just a narrow calculational capacity. Quandary ethics, whether operating through deduction from principles or by application of the MEU principle, does not—cannot—appreciate the nature and momentousness of their task or all that it takes to grapple with their task. Re-reasoning ethics is about getting these matters right. But what does it mean to “re-reason” ethics, and how can ethics be “rereasoned”? Morality has to be rational. If morality is not rational (and not, we suppose, imposed by divine force or equivalent), then it becomes simply a matter of personal opinions or idiosyncratic preferences. Applying a priori moral principles or rules to the facts of problems to derive conclusions about what ought to be done can claim to be rational because it relies on the certain truth of the principles and utilizes the paradigmatic rationality of deductive logic and mathematics. Yet the deliberation of the women after genetic counseling, as we have seen, involves much more than this: it concerns no less than reimagining everyone and everything involved; it is complex, contextual, and riven with vagueness and uncertainty. Is their decision making rational? Not in the terms of the logic-based approach with its formal reasoning that renders deduction and mathematics rational. But what else can rationality be? Pincoffs recognizes the philosophical incredulity that greets any proposal to look elsewhere for rationality: “To abandon the search for rationally defensible rules and principles is to abandon moral philosophy” (1986: 14). Any other conception of rationality becomes (literally) unthinkable. Kurt Baier provides a crisp moral illustration of this philosophical dogma in his explanation of “why we can often tell immediately (as if by a moral sense or by intuition) whether a given course of action is right or wrong”: “We do so by our reason, by subsuming a given course of action under a given principle” (1958: vii–viii). The reduction of rationality to formal reasoning is so pervasive that even Lippman and Fraser echo it. They note that data from descriptive studies

18 

Chapter 1

of “real-life” decision making indicate that when people make complex choices, for example, selecting a spouse or a career, they more often than not violate the normative model that is regarded as appropriate. Lippman and Fraser attribute this quirky “real-life” decision making not to the abandonment of rules but to the invocation of unauthorized rules: “The choices made are neither capricious nor random, but they follow rules other than maximizing expected utility” (1979d: 327–328). Their explanation assumes that decisions can be rational only if they are rule-governed; decisions that are not rule-governed are arbitrary or capricious. Because the deliberation of the women they studied goes well beyond rule-bound reasoning, the rationality of their decision making cannot be understood in the conventional way. Lippman and Fraser concede that: “the notion of rationality may not only be inappropriate, but may also limit our understanding of counselee’s behavior and the influence of genetic counseling on this behaviour” (1979d: 325). And they go further: “Parents’ perceptions of their situation as uncertain in a sense obliges them to follow what might be called ‘a-rational’ methods in resolving their choice problems. Rationality becomes irrelevant” (1979d: 330). That drastic, regrettable conclusion follows only if rationality is restricted to formal reason—to the exercise of deductively applying principles and rules. But why accept the dichotomy, “either rule-bound and rational or not rule-bound and a-/irrational”? Many intelligent problem-solving activities are not rule-bound. Ironically, the construction of logic proofs itself, for example, is provably not a rule-bound process for nondecidable logics (most of them), yet it is a thoroughly skilled, intelligent activity. Creating a new logic still more so. Are mathematicians and philosophers who devise proofs or create logics acting a-/irrationally? Supposing that would divorce rationality from intelligence, thereby gutting both. Accepting that dichotomy commits what Richard Wasserstrom, writing about judicial decision making, insightfully calls “the irrationalist fallacy”: The question of whether an argument is formally valid is confused with the question of whether there can be good reasons for believing a proposition to be true or false. It is recognized quite correctly that the canons of formal logic have a decidedly limited applicability. It is inferred quite incorrectly from this that all the questions which cannot be settled by appealing to formal logic cannot be settled in any manner which could be called “reasonable” or “logical.” An appeal to formal logic is equated with an appeal to criteria of rationality or reasonableness, and it is concluded that because the courts cannot have used formal logic to select or formulate legal premises, the courts cannot have appealed to any rational or objective criteria when engaged in these undertakings.… It may not

Introduction 19

make much sense to describe the judicial decision process as a completely deductive one. But it makes even less sense to insist that for this reason courts could not (and should not) employ a procedure or a set of procedures that permits of some kind of reasoned justification for the judicial decisions reached by those courts. (1961: 23–24)

Likewise, it makes no sense to regard the decision making of the women whom Lippman and Fraser studied as a-/irrational. Their situations are saturated with uncertainty and ambiguity. They must construct a process that allows them to manage that uncertainty and that ambiguity, to frame their problems and the information they receive in ways that are meaningful to them, and to arrive at a conclusion about what they should do. They accomplish all that, and the outcomes of the processes they devise display an immeasurably richer understanding of their situation and of themselves and their family in that situation than what they had at the outset and that any formal calculation could provide. At the same time, they dramatically improve the practical adequacy of their assessments. Moreover, their deliberation confers an enhanced capacity to tackle other such problems in the future. They have learned, and while doing so they have learned how to learn about resolving such deep problems. The processes these women use to structure their problems and to work their way through them are dynamic, creative, and contextual. And a bewildering array of diverse considerations are rooted sensitively and systematically in those processes, thereby belying charges of arbitrariness and capriciousness. Incorporating what these women have to teach into the practices of genetic counseling would improve both the intelligence and the moral quality of counseling processes. At the very least the deliberation of these women is a model of powerfully intelligent, nonformal problem solving. In fact, to deny rationality to their problem-solving process would be to deny it to all the fundamental dimensions of our lives. How to throw a good party, settle a political disagreement, develop a new technology, solve a crime, design a unique piece of jewelry or a new kind of bridge, judge a legal case, become a saint, research a new scientific domain, or constructively parent a child? They are all problems that, like the problems of these women, are complex, multifaceted, multivalued, highly contextual, and mutually interactive, and they all begin ill-defined in form, method, and boundary. Let us call them deep problems. They are just as impenetrable to narrow inferential/calculational rationality as are the women’s problems. And, as we shall see in chapters 3 and 4, scientists use this very kind of process when approaching deep research questions. Thus to call these women a- or irrational is tantamount to denying rationality to virtually the entire substance of intelligent life.

20 

Chapter 1

So what are we to say about the rationality of these women? Essentially there are only three possible responses. The women can be held to be: (A) intelligent, but a- or irrational because they do not use deduction to apply formal principles, for example, MEU; (B) a- or irrational, so unintelligent; or (C) intelligent, so rational. Option B precludes identifying intelligence with deep problem solving, simply to preserve the undefended assumption that rationality must be deduction from principles or rules. Ironically this would cut intelligence adrift from all but the most superficial problem solving. Option A sunders rationality from intelligence. We see no practical or theoretical benefit in doing that. To the contrary. The current sharp divisions among conceptions of intentionality, rationality, and intelligence— namely, intentionality modeled on linguistic reference, a late-evolved trait in human beings; rationality modeled on simple deduction from principles or rules; and intelligence modeled on formal, for example, computational, problem solving—offer an implausible biological picture of the evolution of isolated mental tricks. As opposed to this account, the model of deep problem solving represented by the women’s methods, as we shall see, a model widely applicable in science, law, design, and elsewhere, supports a more plausible evolutionary model that understands intentionality and rationality as but two complementary facets of a single, developing problem-solving and decision-making capacity. Thus we conclude that option C thoroughly deserves our allegiance.6 For all these reasons, ethics needs to be re-reasoned. To that end, the regnant conception of rationality needs to be expanded beyond the narrow equation of rationality with formal reasoning to account for the rationality of the deliberation of the women who received genetic counseling, along with much of the deliberation that suffuses our lives. Developing that broader conception of rationality and articulating its wide-ranging consequences for ethics are the complementary projects of this book. 1.4  Four Lessons Four key lessons emerge from the case presented in this chapter. The first and most obvious lesson is that the dominant conception of rational procedure constrains roles and outcomes in problem-solving or decision-making situations. The genetic counselors’ implicit understanding of and commitment to formal rational decision making clearly constrained both their counseling role (to just providing empirical outcomes and probability data) and the women’s role (to just combining that empirical information with their personal evaluations of outcomes and then cost-benefit

Introduction 21

maximizing to reach a decision). Constraining behavior is inherent to any account of normativity, so constraint per se is not the issue. Rather, what this example highlights is that if we judge that the decision-making approach to which a normative account constrains us is inadequate, as these women did, the issue that properly arises is that of whether there is something wrong with that normative account (and not with the women). In short, second, we can reasonably test the adequacy of a normative account by evaluating the particular decision-making approach it requires against our knowledge of the alternative problem-solving processes we are able to undertake. And if there is an alternative approach that is judged superior, as happened with these women, then, third, it can be rational for the redesign to reciprocally include revising the nature and role of the rational procedure within which the superior approach is embedded. Then the issue that directly arises is what alternative normative account might better include within its obligatory constraints the decision-making approach considered most valuable. In their decision making these women identified and focused on much larger issues than what logical procedures address. They were concerned about what the design of a flourishing life meant for them in all their present circumstances. So we conclude, fourth, that we will not be able to pass to the primary concern of designing a life proper without reconsidering the nature of rational ethical process, that is, of an appropriate problem-solving/decisionmaking approach. Twin themes that structure the entire book emerge from these lessons. (1) We want to develop an alternative, more expansive conception of reason (alternative, that is, to the conception of reason as formal calculus) that is sufficiently rich to encompass all the procedures through which we humans can intelligently design our lives and our world to promote human flourishing in all its aspects. (This alternative account should describe a process that is itself learnable and improvable as we go about those very tasks.) (2) We aim to develop a sufficiently rich account of the proper scope and methods of rational ethics in this larger life-design task and show how these methods can lead to approaches and outcomes that are superior to those yielded by the methods of orthodox moral philosophy, indeed throughout the book to illuminate important cases whose resolutions are neglected because they are not representable within the logic-based approach.

2  The Difficulties of Applied Ethics

C

Bioethics, like its counterparts in the ethics of law, business, engineering, journalism, and other domains, generally has been regarded as applied ethics. In the first edition of their influential text, Principles of Biomedical Ethics, Tom Beauchamp and James Childress promulgated this view by characterizing bioethics as “the application of general ethical theories, principles, and rules to problems of therapeutic practice, health care delivery, and medical and biological research” (1979: vi–vii). Over six subsequent editions, Beauchamp and Childress have refined and embellished their method in response to a panoply of criticisms, while bioethics, despite all these changes, continues to be widely portrayed and practiced as a matter of applying “general ethical theories, principles, and rules” to concrete moral problems. The persistence of the applied ethics model is easy enough to understand. In practice morality shades seamlessly into etiquette, law, or practical convenience, and the attempt to demarcate the distinctively moral becomes problematic. Moral norms might be differentiated from rules of etiquette in terms of their greater importance, and from legal rules in terms of their informal origins and the kinds of social sanctions imposed when they are violated, but these distinctions are anything but sharp and intrinsic.The only alternative to justify norms as genuinely and distinctively moral, it is widely accepted, is to organize them into a distinctive and consistent system of principles or rules to form a moral theory, perhaps through axiomatization. The only sure way to accomplish this theory construction— and the one typically adopted—is to appeal to a priori insight. Then, right actions can be logically deduced by applying the theory to the empirical features and circumstances of situations, thereby rendering practical morality a rational, objective enterprise. This conception of morality is not only pleasingly tidy, it also makes morality an autonomous discipline.1 The conception of bioethics as applied ethics incorporates this familiar philosophical account of what morality is and how moral justification

T

24 

Chapter 2

proceeds. Bioethics is, in this view, a norm-constituted theory, and in what has become the canonical understanding of bioethics, that theory comprises Beauchamp and Childress’s four fundamental principles: a principle of nonmaleficence, a principle of beneficence, a principle of autonomy, and a principle of justice.2 Bioethical decision making is a matter of applying these principles and subordinate norms derived from them, such as a principle of informed consent, to moral quandaries in health care. Despite the hegemony of this approach, conceiving morality in this way is, as we saw in chapter 1, far too simple. The methodological ferment in bioethics that followed the publication of the first edition of Beauchamp and Childress’s Principles of Biomedical Ethics and that continues today testifies to the weaknesses and limitations of applied ethics thus understood. This chapter uses vignettes to depict core issues in bioethics that illustrate those failings in a practical, concrete manner. The vignettes display the profound limitations and deep problems that characterize any such model of applied ethics and the moral philosophy from which it is derived. 2.1  Indeterminacy: What Does a Norm Mean? To apply a norm—a principle or a rule—one must know what it means. For norms to be norms, that is, to be directives applicable to an array of cases and not just commands for specific cases, they must be appropriately general. Moral principles and rules assign a normative status—for example, permissible, obligatory, or forbidden—to general classes or types of act. But then the relevant classes or types of act to which they are applicable must be specified; that is, criteria must be provided for identifying the particular acts that fall within the ambit of a class or type to which a norm applies. For instance, characterizing assault as including any plunging of a knife into a living human body will not do since it includes medical surgery to save lives and plastic surgery to save face; but it ought to include a surgeon’s deliberate, unnecessary surgery performed in revenge. The world is richly complex, dynamic, and protean; the general terms of a natural language are vague; and we are finite and fallible. We cannot understand everything about the world, and we cannot anticipate the multifarious ways in which our aims, goals, and values will interact and change and will impinge upon the world. So classes of acts never can be rigorously, precisely, and exhaustively defined, and criteria for identifying the acts that those classes comprise never can be correspondingly detailed. Problems of classification inevitably arise, and those problems create indeterminacy in the application of moral rules and principles. Consequently, judgments

The Difficulties of Applied Ethics 

25

about when and how to apply these norms have to be made. The problem of classification plagues all of applied bioethics but is particularly compelling with respect to two of its most developed and contentious topics. 2.1.1  The Meaning of “Euthanasia” Discussions of the moral and legal permissibility of euthanasia can quickly founder on uncertainty and disagreements about the meaning of the term. Animals are “euthanized” with a lethal injection. Must euthanasia for persons be similarly “active,” or can it be “passive”? If there is a difference between active and passive euthanasia, is that because active euthanasia is killing, whereas passive euthanasia is allowing to die? Also, conceptual unclarity, and its attendant moral confusion, create practical problems of classification. Is the withdrawal of life-sustaining treatment, such as a mechanical ventilator, an instance of euthanasia? Is the withholding of life-sustaining treatment, such as antibiotics for pneumonia in a severely demented resident of a nursing home, an instance of euthanasia? The first vignette poses another classification problem, one that concerns the potential use rather than nonuse or discontinuation of an intervention. Vignette 2.1  Ms. B Ms. B, who is 30 years old and single, was told by her doctor on Christmas Eve that she had an ovarian mass that might be cancer. Supported by her sister, her only sibling, and her mother and father, she struggled with the uncertainty, anxiety, and fear of this possibility throughout the holidays. Afterward she was diagnosed as having a serious form of ovarian cancer that required immediate surgery and chemotherapy. Over the next two years Ms. B endured bouts of radiation therapy, more chemotherapy, and more surgery, intermingled with times when she could feel her life still was her own. At the end of those two years, Ms. B had a colostomy. She was expected to live for only a few more months. Soon thereafter, Ms. B became bedridden at home and clearly was dying. She asked Dr. M, her palliative care physician, whether he remembered all the talks they had had over the past few months about how afraid she was to die struggling for breath. Ms. B reiterated that her physical pain was well controlled, but she stressed that her breathing was getting harder and harder, which scared her and made her breathing difficulties even worse. Dr. M assured her that he did remember, that he would not abandon her, and that he would not forget her wishes. Dr. M promised Ms. B that he would sedate her if her breathing difficulties could not be controlled. Ms. B assured Dr. M that the morphine helped her a lot, but she said she knew she was taking more and more to control her symptoms, and she felt she would need to be sedated soon. She added that she knew she was dying and she was not afraid to die. She knew she was going to a better place, and she just wanted to sleep and be at peace before getting there. She no longer wanted

26 

Chapter 2

any food, and she was taking only sips of water and occasionally a small piece of popsicle. One day a good family friend visited Ms. B, and they discussed her fear of suffocating and what would be done if that were to occur. The friend commented that he understood the plan and that he was a strong supporter of euthanasia. Ms. B was startled and nonplussed, and she became anxious and upset. After her friend left and Ms. B told her family what he had said, they became distraught and called Dr. M. The family explained that they were very religious and that although they appreciated everything medicine could offer, they did not want Dr. M to “euthanize” Ms. B if her breathing became difficult. Dr. M was shocked by the suggestion.3

Terminal sedation is used in palliative care4 to relieve intolerable pain or suffering that cannot be controlled. Ms. B’s pain is well controlled. Ms. B is not afraid of death. What Ms. B is afraid of is how she might die. She fears that when her breathing difficulties worsen, she will start gasping for air and feel that she is suffocating. That fear causes her anguish and her suffering. To avoid that suffering, Ms. B wants to be sedated to unconsciousness. She is eating only small bites of a popsicle and drinking only small sips of water, so once she is unconscious, all life-sustaining interventions will be withheld, including artificial nutrition and hydration. She will die of dehydration, starvation, or a complication or deterioration of her underlying terminal condition. Ms. B, her family, and Dr. M agree about the governing norms: relief of suffering is morally obligatory, euthanasia is morally prohibited. This would seem to make the situation a straightforward case of applying these rules in the standard applied bioethics model. But there remains the difficulty of how to apply the euthanasia norm to their situation. Is terminal sedation an instance of euthanasia? Different answers to that question create moral turmoil. The family friend, who strongly endorses euthanasia, thinks it clearly is. Dr. M, whose shock shows that he strongly rejects euthanasia, thinks it clearly is not. Ms. B and her family, who also strongly reject euthanasia, are upset that it might be. If that matter of classification were settled, however, a solution to their moral quandary, according to the applied ethics model of bioethics, would follow swiftly and surely. A statement about whether terminal sedation is or is not euthanasia would be subsumed under the governing norm, and a conclusion about whether terminal sedation is morally permissible for Ms. B would be logically deduced. But how is that crucial matter to be settled? Is what appears to be a separate, well-defined problem of semantic classification—what does “euthanasia” mean?—really so delimited and independent of moral considerations?

The Difficulties of Applied Ethics 

27

Notice, for instance, that the question already has changed. The question no longer is about only terminal sedation, but terminal sedation in combination with the withholding of all interventions that might prolong Ms. B’s life. Does this more complex plan make her envisaged death a form of euthanasia? The term “euthanasia” comes from Greek, where, taken literally, it means a good death (eu = good + thanatos = death). In that sense the mode of death proposed for Ms. B is, if not a good death, at least a better death than a suffocating death. Over two millennia, however, and more recently with all the technological advances in medicine that enable life to be extended well beyond previous limits, “euthanasia” has acquired complicated moral, legal, and social meanings. The meaning of “euthanasia” has become implicated in and shaped by the moral controversies it creates. Two prominent moral positions that address issues about when and how life may be ended have particularly influenced the understanding of euthanasia. One is the distinction between active and passive measures, and the other is the doctrine of double effect. The difference between an active mode of hastening or causing death and a passive mode of hastening or causing death is, for many people, morally crucial. Active measures, such as administering a lethal injection at the request of a patient in the case of voluntary active euthanasia (VAE), are deemed morally impermissible, whereas passive measures, such as withdrawing or withholding life-sustaining therapies, are deemed morally permissible. The putative moral difference turns on what appears to be substantially different methods of causation. Whereas in VAE a doctor directly causes the death of a patient by administering an injection, in withholding or withdrawing life-prolonging interventions it is the patient’s underlying disease that causes death. In the former, human intervention causes death; in the latter, nature causes death. That distinction can be difficult to sustain practically, however, because doctors take a series of active steps in withdrawing a patient from a mechanical ventilator. The distinction also is difficult to sustain theoretically. Patients who are terminally ill can end their lives by deciding to voluntarily stop eating and drinking (VSED). As Timothy Quill, Bernard Lo, and Dan Brock describe it, “With VSED, a patient who is otherwise physically capable of taking nourishment makes an active decision to discontinue all oral intake and then is gradually ‘allowed to die,’ primarily of dehydration or some intervening complication” (1997: 2099). VSED can be regarded as a passive measure, along the lines of withholding or withdrawing life-sustaining interventions, in which the cause of the patient’s death is an underlying disease. But as Quill, Lo, and Brock object, “the notion that VSED is passively ‘letting nature take its course’ is

28 

Chapter 2

unpersuasive, because patients with no underlying disease would also die if they stopped eating and drinking. Death is more a result of the patient’s will and resolve than an inevitable consequence of his disease” (2102). Might the active-passive distinction nevertheless allay the worries that Ms. B and her family have about terminal sedation? Terminal sedation could be regarded as morally permissible because “the administration of sedation does not directly cause the patient’s death and because the withholding of artificial feedings and fluids is commonly considered passively allowing the patient to die” (Quill, Lo, and Brock 1997: 2102). Doctors and nurses, however, probably would consider the ongoing administration of the drugs needed to sedate a patient to unconsciousness to be an active process; moreover, the suggestion that, in terminal sedation, nature is taking its course is questionable because the patient often dies of dehydration from the withholding of fluids, not of the underlying disease (2102). What about the doctrine of double effect? The doctrine of double effect is a technique for morally evaluating courses of action that have two consequences, one that is desirable and another that is undesirable. The suggestion that terminal sedation is a kind of euthanasia might have prompted Ms. B and her family to appreciate that aggressive sedation could depress Ms. B’s respiration and thereby cause her to die sooner than she otherwise would without it. In that event, terminal sedation would have a double effect: the desirable effect of preventing suffering and the undesirable effect of hastening death. Would the latter make it morally wrong? A central component of the doctrine of double effect is the intention with which one acts. For a course of action to be morally permissible, the desirable outcome must be intended, and the undesirable outcome must be unintended, even though it is foreseen, perhaps with a high degree of probability. It could be argued that the intent of administering terminal sedation is relief of suffering and that the acceleration of death, although foreseeable and highly probable, is unintended. But terminal sedation is accompanied by the withholding of life-sustaining measures, including artificial nutrition and hydration. With both, the dying process usually takes days, if not weeks; without withholding life-extending measures, the process would be substantially prolonged. When withholding artificial nutrition and hydration is added to the administration of terminal sedation, however, it is difficult to claim that only relief of suffering is intended, as Quill, Lo, and Brock explain: Although the overarching intention of the sedation is to relieve the patient’s suffering, the additional step of withholding fluids and nutrition is not needed to relieve pain, but is typically taken to hasten the patient’s wished-for death. In contrast, when patients are similarly sedated to treat conditions like status epilep-

The Difficulties of Applied Ethics 

29

ticus, therapies such as fluids and mechanical ventilation are continued with the goal of prolonging life. (1997: 2101)

Were withholding fluids and nutrition to be added to Ms. B’s plan for terminal sedation, as a reasonable extension of her current behavior and as a reasonable thing to do in any event, the intent of that combination would be not just to end her suffering but also to hasten her death as a means to that end. So the friend of the family would be right. Although the moral distress of Ms. B and her family is caused by the suggestion that terminal sedation is a form of euthanasia, asking whether terminal sedation is an instance of euthanasia as a preliminary to applying a moral norm does not help them grasp their moral situation more clearly or address their fears and worries. Like many general terms, especially those described as “essentially contested” (Gallie 1956), “euthanasia” does not have a rigorous, precise, unequivocal meaning. Deciding whether terminal sedation is a form of euthanasia cannot be a straightforward, determinate matter of applying a concept, as happens, for example, when a class of elementary school students is given a sheet of geometrical figures and told to color the triangles. The meaning of “euthanasia” is dynamic, shifting over time and across circumstances, and it is influenced by moral, legal, and social controversies, most recently, by ones about withholding and withdrawing life-sustaining treatment and legalizing physician-assisted suicide and voluntary active euthanasia. There is another problem. The moral perplexity and distress created by Ms. B’s situation focuses on the notion of euthanasia because it is assumed that the relevant governing norm concerns the moral status of euthanasia. That assumption can be questioned. Quill, Lo, and Brock, for example, reject an approach to determining the moral status of voluntarily stopping eating and drinking, terminal sedation, physician-assisted suicide, and voluntary active euthanasia that is predicated on applying the active-passive distinction or the doctrine of double effect. For them, what is morally important, with respect to competent patients, is the ethical principle of respect for autonomy. None of the four ways of ending life would be morally permissible, they emphasize, without the voluntary and informed consent of the patient. Ms. B is competent. Ms. B has discussed her fears with Dr. M, and they have come to an agreement about how to treat her impending breathing difficulties. She has made a voluntary, informed, considered decision about terminal sedation, and she has sought Dr. M’s reassurance that he will keep his word. Respect for the autonomy of Ms. B requires that Dr. M provide terminal sedation.

30 

Chapter 2

This moral alternative rejects the norm that initially is assumed to govern and structure Ms. B’s plight and reverses the applied judgment. So what is the proper moral approach here? Is a decision about terminal sedation for Ms. B a matter of the moral permissibility of euthanasia or a matter of respect for autonomy? On the basis of their analysis, Quill, Lo, and Brock conclude that “TS [terminal sedation] and VSED [voluntary stopping eating and drinking] are more complex and less easily distinguished ethically from PAS [physician-assisted suicide] and VAE [voluntary active euthanasia] than proponents [of VSED and TS] seem to realize” (1997: 2101). Trying to identify relevant similarities and dissimilarities among those four ways of ending life and then basing moral assessments on the results of those comparisons is, in their view, misguided. For them the common and central moral issue posed by all four, with respect to competent patients, is whether the patient’s decision is voluntary and informed. Their approach takes us back to where we began. Like “euthanasia,” “autonomy” is a general term. What does autonomy mean, and how are particular decisions of particular patients about particular problems to be classified as autonomous or nonautonomous? 2.1.2  The Meaning of “Autonomy” In the second vignette respect for autonomy becomes the focus of life-anddeath decision making. Vignette 2.2  Ms. F Ms. F, who is 28 years old and has cystic fibrosis, had a lung transplant two years ago. She began experiencing problems with her new lung several months after she was discharged from the hospital, and she subsequently had a heart-lung transplant, after which she was able to return home. A month later she had to be readmitted to the hospital because the second heart-lung transplant also was unsuccessful. The lung transplant team rejected the possibility of a third lung transplant because it was considered to be medically futile. Ms. F was encouraged to consider a hospice program in which, she was assured, her medical problems would be treated and palliative care would be provided when necessary. Even though transplantation of a third lung in the same patient is rare, the conclusion that another lung transplant for Ms. F would be futile was not unanimous. A physician on the transplant team, who was about to join the staff at another hospital to direct their new lung transplant center, thought that Ms. F might be able to survive another transplant, and he communicated his opinion to her. If she wanted to be considered for a third lung transplant she could be transferred to the other hospital. Ms. F faced an agonizing decision. She knew that if she remained where she was, death would be imminent. She also knew that if she wanted another chance, she would have to leave the hospital where

The Difficulties of Applied Ethics 

31

the lung transplant team had become, over the previous two years, a surrogate family to her.5

The centerpiece of contemporary bioethics is a principle of autonomy that authorizes patients to make decisions about their health care in terms of their own beliefs and values. Cognizant of this principle, the hospital staff request an ethics consultation for Ms. F. They are willing to respect whatever decision Ms. F makes, provided they can be confident that her decision is genuinely autonomous. How is a clinical ethicist to make that determination? What does it mean for a decision to be autonomous? Bruce Miller has neatly distinguished four components of autonomy: autonomy as free action, authenticity, effective deliberation, and moral reflection (1981). An action is free when it proceeds from a decision that is intentional and voluntary, that is, not the result of coercion, duress, or undue influence; authentic when it is consistent with a person’s attitudes, values, and goals, that is, “in character” for that person; the result of effective deliberation when the person deciding what to do is aware of the courses of action that are available and the likely consequences of these alternatives, evaluates them, and makes a choice based on that evaluation; the product of moral reflection when a person scrutinizes the values that have been inculcated in him or her, accepts a set of values as his or her own based on that critical examination, and then proceeds to make decisions based on those values. Given Miller’s analysis, the clinical ethicist first must determine whether Ms. F’s decision is autonomous in each of the four senses. Free action: Whatever Ms. F decides would be intentional, but would it be voluntary? Ms. F is torn not just between her options, but between her two families. Her parents and her sister support the decision of the lung transplant team, but her boyfriend and his family want her to try again. There is pressure in both directions, so no matter what Ms. F decides, she will be doing something that somebody close and important to her believes is wrong. Do the competing pressures somehow cancel out, or should they simply be accepted as a fact of life? If the latter, what distinguishes troubling yet morally benign pressures from pressures that render decisions unfree—pressures that constitute undue influence or duress? In this regard, consider how important the lung transplant team has become to Ms. F. Members of the team have been caring for her for months at a time over a period of two years. Ms. F understands how dependent she is on them, and she recognizes how she resists the authority they have and the power they wield over her. One of her journal entries reads: “I woke up in a horrible mood. Sick of being here

32 

Chapter 2

in the same place, no patience for anything, being a whiner. All it does is make things harder. Jim is my nurse. I know I must piss him off. I got a big dose of Morphine and that just about knocked me out— so now I’m laying dizzy and paying for it. I bring this on myself I think” (Marshall 2001: 142). Ms. F’s emotions are roiling, and her moods can be dark. She recognizes her intemperate reactions, and she knows her bad moods have consequences: “a big dose of Morphine” in this case. Given her dependencies and vulnerabilities, there is strong pressure on Ms. F to be compliant, to be a “good” patient. Is that kind of pressure morally innocuous, or does it raise concerns about the autonomy of her behavior and her decisions? Effective deliberation: Ms. F agonizes over her decision, strives to understand her predicament, and seeks guidance in her religious convictions. Does that qualify as deliberation, and if so, is it effective deliberation? The model of effective deliberation in bioethics is exemplified by the informed consent procedure. To obtain valid consent for treatment from a patient, a health care professional must inform a patient about the nature of the problem, alternative forms of treatment, the likely benefits and risks of those alternatives, and the likely outcome of doing nothing. The patient then is supposed to assess the consequences of these options in terms of his or her own values and make a decision on that basis. To facilitate this kind of deliberation, the clinical ethicist could construct a decision tree and guide Ms. F through a detailed consequentialist process of reasoning. Chapter 1 already has provided a telling lesson about the moral traps in this approach. In that case effective deliberation began by rejecting the whole apparatus of benefit-cost maximization. And here it turns out that such an exercise might not be meaningful to Ms. F because she too does not structure her problem in that way. Ms. F’s thinking is more metaphorical. During the consultation, the clinical ethicist reads a poem Ms. F has written in which she compares her body to a bicycle with stripped gears. A portion of the poem goes: “… but your bike is starting to fall apart, the gears stripped, leaving you with the exact one you don’t need. … They fix the gears and tell you go on … keep going it’s gonna get easier, something’s wrong though cause you fall again, hard and this time the gears can’t be fixed. … A lot of people say never give up. Others tell you to do as … best you can. How do you learn to accept? Or do you fight ‘till the absolute end’?” (Marshall 2001: 144). Bikes break and are taken to mechanics to be fixed. Bodies break and are taken to surgeons to be fixed. Indeed, there is a surgeon ready and willing to try to repair Ms. F’s broken body. What should the clinical ethicist make of Ms. F’s metaphor? In particular, should the clinical ethicist challenge

The Difficulties of Applied Ethics 

33

the mechanistic, reductionistic image of persons and their bodies that it conveys and point out the kind of bias this assumption introduces into Ms. F’s decision making? Ms. F is thinking about her situation and trying to work her way through her problem, but is she engaging in the kind of deliberation required for her decision to be autonomous? What should the clinical ethicist conclude if her process of deliberation does not match the consequentialist reasoning presupposed by the exercise of obtaining informed consent?6 Authenticity and moral reflection: Assessment in these two categories is even more difficult because it requires historical and personal information about the social, psychological, emotional, and spiritual dimensions of Ms. F’s life. Would a clinical ethicist be able to establish the trust and intimacy required just to broach such topics? Would Ms. F be willing to allow her family and friends to talk about private aspects of her life with a clinical ethicist? Would a clinical ethicist have the training and the skills needed for that kind of interviewing? Although the senses of free action and effective deliberation convey the kind of control autonomy protects, the senses of authenticity and moral reflection convey the deeper meaning of autonomy that gives that kind of control its moral importance. Satisfaction of the former two senses establishes the existence of control, but the latter two senses are necessary to understand whether and how a person wishes to exercise control. Were the clinical ethicist able to sort out these indeterminacies, another problem could arise. Ms. F’s decision might be autonomous in some senses but not autonomous in other senses. Are all four senses required for a decision to be autonomous? Do mere numbers count, or are some senses more important than others? How, in other words, is the clinical ethicist to reach a conclusion about whether Ms. F’s decision is, all things considered, genuinely autonomous? Miller recognizes that “what sense of autonomy is required to respect a particular refusal of treatment is a complex question” (1981: 27). Two considerations that Miller cites as relevant to answering that question are the nature of the relationship between the person making the decision and the person examining the autonomy of the decision, and the seriousness of the decision. What kind of relationship could Ms. F have with a clinical ethicist whom she did not ask to see, of whose existence she probably had no knowledge, and whose line of work undoubtedly would be baffling to her? And even if a close relationship between Ms. F and the clinical ethicist were to emerge quickly, how would that be relevant to assessing the autonomy of Ms. F’s eventual decision? Perhaps the relevant relationship is

34 

Chapter 2

not between Ms. F and the clinical ethicist, though, but between Ms. F and the lung transplant team. The clinical ethicist is merely an agent of members of the team with whom Ms. F has established close relationships. The relevance of close relationships to determining the autonomy of decisions remains unclear, however. Are there different tests for relationships with strangers and relationships with intimates, with lay persons and professionals? If so, why and how do those tests differ? Is a bioethicist determining autonomy like a consulting psychiatrist determining competency, or is autonomy even more elusive? Relationships are morally important, but how they are relevant to applying the principle of respect for autonomy is not clear. Miller’s other concern about the seriousness of the decision is straightforwardly satisfied, though. Ms. F is making one of the most important decisions she ever will have to make. But what does that mean for an assessment of the autonomy of her decision? Does it mean that the criteria for the four senses of autonomy should be applied more rigorously and stringently? Is Ms. F’s decision making being scrutinized because the lung transplant team is worried that she might make a mistake? Indeed, in Ms. F’s circumstances does it make sense to talk about a “correct” decision? Miller’s two considerations are important because they recognize the moral relevance of the particularity of persons and relationships and the details of situations and settings. But contextualization raises serious problems for applied bioethics. When interpretations of the various senses of autonomy and assignments of priorities to those multiple senses become context-dependent, the claim that moral decision making is a matter of applying rules is hard to sustain because there would have to be secondorder rules for interpreting the first-order rules about autonomy, in particular for determining which features of persons, relationships, situations, and settings are morally relevant. There also would have to be second-order rules for resolving conflicts between competing interpretations of the notion of autonomy. But then those second-order rules could need interpretation, and conflicts between those interpretations would have to be resolved, so there would have to be third-order rules to govern those decisions. Not only would an infinite regress of rules loom, but our moral insight has somehow to extend to all these new and increasingly abstract rules. Does it really? We suggest not, as indicated by the nearly universal absence of plausible such rules. Conversely, were such rules somehow conjured, they would become so numerous and so complicated that applied bioethics, even if it were still theoretically tractable, would become practically unmanageable.

The Difficulties of Applied Ethics 

35

Moreover, like the first vignette, this one raises the issue of how the governing norm is determined. Is Ms. F’s problem one of questionable autonomy? In an early meeting with the clinical ethicist, Ms. F asked, probably rhetorically, “What is my choice? To live like this for the rest of my life [while gesturing to the monitors and tubes in her hospital room]? Or to try it again and see what happens? Is there a choice?” (Marshall 2001: 146). If Ms. F has no real choice, why the moral preoccupation with respecting her autonomy? Ms. F is trapped, and one has to have enormous sympathy for her, for how she agonizes, and for whatever she decides. Given her plight, how can a clinical ethicist wielding the limited resources of applied bioethics determine whether her decision is genuinely autonomous? That the ethics consultation focuses on the autonomy of her decision making is an artifact of the prevailing applied-ethics model of bioethics, not a perceptive, sensitive response to her dilemma. The job of the clinical ethicist is to help Ms. F, not just the lung transplant team. Recalling the genetic counseling case in chapter 1, the clinical ethicist might be able to do that effectively, not by focusing on value maximization, but by listening to and comforting her, by sensitively assisting her to explore her own perceptions and feelings about her situation, and by supporting her. Yet applied bioethics offers no such alternative. It is no surprise, therefore, that the clinical ethicist can do so little to help Ms. F make her agonizing decision about whether to try a third transplant.7 2.2  Conflicts: Which Norm Prevails? When a moral theory contains more than one principle, conflicts between those principles inevitably arise, as, for instance, the conflict between relieving suffering and avoiding euthanasia for Ms. B in vignette 2.1. In contemporary bioethics the familiar issue of paternalism results from a conflict between a principle of respect for autonomy and a principle of nonmaleficence. After careful deliberation a competent patient might, for example, refuse life-sustaining treatment that a physician believes is likely to provide the patient with a reasonable quality of life. Should the physician respect the patient’s autonomous decision or try to prevent what he or she regards as avoidable harm to the patient? The latter alternative is disparaged as being paternalistic because the physician would be assuming that he or she knows, better than the patient does, what is in the patient’s best interest. Respect for autonomy has, at least in theory, become firmly ensconced in contemporary bioethics, even if that respect eventuates in harm that could have been averted.

36 

Chapter 2

Conflicts about the use of life-sustaining treatment can be even more agonizing, however, when the life at stake is that of an infant, as in the following vignette. Vignette 2.3  K’aila In the weeks after his birth at home, K’aila nursed well, grew, and gained weight. His skin and the whites of his eyes remained a little yellow following an episode of newborn jaundice, but K’aila’s doctor was not concerned. When K’aila was still yellow almost three months later, however, he was examined in a hospital by a pediatrician. Preliminary tests revealed abnormal liver function, but additional tests did not identify the cause of the liver dysfunction. The diagnosis subsequently was narrowed to either biliary atresia or giant-cell hepatitis. There is no treatment or cure for either. K’aila’s mother and father were told that the only way to prolong K’aila’s life, as his condition deteriorated, would be to perform a liver transplant. K’aila’s pediatrician explained that transplantation is the standard, accepted treatment for end-stage liver disease in North America. He cited survival rates that approached 80% to 85% with improved techniques and the use of the drug cyclosporine to prevent organ rejection. K’aila’s mother and father are Native Americans who from the outset had deep misgivings about a transplant. They struggled with their decision and prayed and meditated for months. They also acquired more information about transplantation. They learned that K’aila would have to take immunosuppressants for the rest of his life. They discovered that the estimated 80% to 85% survival rate was high compared with other sources that cited a 70% to 75% survival rate. They realized that these survival rates are for a period of one year and that survivors die at a rate of 1% to 2% each year thereafter, so that the survival rate at five years after a transplant is closer to 60% to 65%. They also discovered that immunosuppressive therapy renders the body vulnerable to infections, the most dangerous of which are viral. Chickenpox, for example, can be fatal to a child on immunosuppressants. And they learned about the side effects of the immunosuppressant drugs, which include progressive damage to the kidneys, cholesterol buildup, hormonal changes, neurologic disturbances, increased hairiness, and a 8 heightened risk of cancer. K’aila’s parents decided to refuse a transplant. K’aila’s pediatrician was uncomfortable with their decision, and because they sensed he would have difficulty providing only supportive care to their son, K’aila’s parents transferred his care to a different pediatrician. The original pediatrician reported K’aila’s parents to the provincial Department of Social Services, which referred them to the Department of Social Services in a neighboring province. Officials of that province applied to the courts for custody of K’aila so that an assessment could be performed and K’aila could be put on the list for a liver transplant without the consent of his parents. The court upheld the right of K’aila’s parents to make a decision that

The Difficulties of Applied Ethics 

37

was within the bounds of conventional medicine and that they believed was in the best interest of their son.

At the most abstract level the disagreement between K’aila’s parents and his pediatrician can be couched in terms of what is in K’aila’s best interest. “Best interest” is, however, an amorphous notion. To K’aila’s pediatrician it means that K’aila should be given a chance to live, even if his life would constantly be at risk and the quality of his life inevitably would be compromised by unending medical problems, complications, and side effects. To K’aila’s parents it means that the suffering K’aila would have to endure is not worth the risky grasp on staying alive that it might allow. Applied bioethics depicts this dilemma as a conflict between principles, one venerating the quality, the other venerating the quantity, of life. Which principle is more important in this situation? How can that be determined? For applied bioethics there would have to be a higher-level norm that governs that choice. But there are many such conflicts among principles, so there would have to be many higher-level norms, and we must expect that they also would conflict, requiring still higher-resolving norms, and so on.9 How, then, does applied ethics determine which principle is right when there is a conflict? Intuitions might be invoked, but, as the vignettes illustrate, there are few such intuitions, and they are weak and/or fuzzy. More common is a resort to “balancing” or “weighing” principles, an approach that, as the Statue of Liberty symbolizes beautifully, turns out to be no more than a metaphor, lacking any substantive correlate of gravity to do the work. Moreover, all such approaches conceal the contextual judgments that are needed to assess how much importance norms have in the unique particular situations to which they are brought. By portraying moral problems as conflicts between principles, applied bioethics fosters the impression that, in the likely absence of a dispositive higher norm, these conflicts either are stalemates or are settled only by appealing to subjective, nonrational considerations external to norms. Moreover, imposing this norm-constituted, norm-governed reasoning on moral problems simplifies and distorts the complex nature of real moral deliberation. K’aila’s mother provides an illuminating and sensitive account of how she struggled to come to a decision: The more I learned, the more integrated my decision became. I kept coming back to the same choice, at several levels, until it became clear to me that I could not live with any other. Had I given in to pressure to accept intellectually an ill-defined statistical chance that K’aila might survive an unknown number of years, with an

38 

Chapter 2

unpredictable number of problems and side effects, I would have had to deny what I knew at other levels. I would have been torn for the rest of my life, knowing that I had made a decision that didn’t feel right inside. (Paulette 1993: 15)

The reflection that was meaningful to K’aila’s mother had nothing to do with numbers and rules. Her deliberation operated at several levels, so she was not thinking just about K’aila but also about how she was thinking. Moral deliberation requires that multilevel complexity and reflexivity: thinking about a problem and at the same time reflecting on that thinking. With respect to the former, for example, how open to be to others who are affected and how spiritual and medical considerations are to be related; with respect to the latter, how one is oriented toward the problem and how sensitive to be to the particularities of the situation. K’aila’s mother reports that “it was in quiet moments with K’aila that I felt most lucid” (Paulette 1993: 16). One of those lucid moments occurred in a hospital ward when an intravenous line in K’aila’s foot was making him uncomfortable: At last I took him out of his crib and brought him to bed with me on a narrow cot. As he snuggled into my arms and began to nurse, he relaxed and fell asleep. I watched him sleeping peacefully, and it became clear to me what I had to do for him. I imagined him like a butterfly, so beautiful, indeed exquisite, and yet so fragile. With my palm open wide, he had alighted on my hand and now rested there a few moments, his wings shimmering in the light. I longed for him to stay forever, but I knew that at any time he might flutter his wings and be gone. The only way I could hang onto him would be to close my hand around him, to entrap him in my grasp. However, I knew that in doing so I would run the real risk of harming him; I might open my fingers again to find his wings broken and crippled, his colors faded and smudged. I knew in that moment that I must keep my palm open and let him take flight when the time came. (Paulette 1993: 16)

The reasoning that engages K’aila’s mother—all of her, not just her intellect— is metaphorical. It is that reasoning that moves her and on the basis of which she decides what to do. In this respect as well, the norm-governed reasoning of applied bioethics ignores the profound contextuality of moral deliberation. Moral deliberation is grounded in background epistemological and metaphysical assumptions that both structure and give meaning to the deliberation. When confronted by the prospect of a transplant for K’aila, his mother and father “had a general and overwhelming sense that it would be an unnatural thing to do” (Paulette 1993: 14). K’aila’s mother explains the metaphysical context of their belief: Deep within me there is knowledge of the intimate relationship between the spirit and the body. One lives on long after the other has fallen away but, in this

The Difficulties of Applied Ethics 

39

brief interval, the two are intimately connected. This knowledge provides the basis for many specific practices and observances in my daily life. One, for example, is the care with which I dispose of my hair when it has been cut. My hair is part of me, associated with my spirit, so I can’t just allow it to be swept into the garbage. So it was that I found it deeply disturbing to consider taking out part of K’aila’s body, given to him by the Creator, and discarding it in favor of part of someone else’s body, associated with a different spirit. Such a radical intervention in his life, which not only would defy the physical laws of nature but also might disrupt his spiritual path, seemed utterly perverse. (Paulette 1993: 14–15)

Associated with these metaphysical views about bodies, spirits, and persons are epistemological views about what can be known and how it can be known. K’aila’s mother articulates those as well: I tried to explain these feelings to the pediatrician, but I’m not sure that he ever understood. These were not merely “religious taboos” or prohibitions that prevented me from making a “rational” choice. These feelings were manifestations of knowledge that was available to my spirit, long before it was ever available to my mind. The human spirit can reach horizons far beyond what the eyes can see and can perceive the path that lies ahead long before the mind can make sense of the signs that mark the trail. (Paulette 1993: 14)

That explanation undoubtedly was hard for the pediatrician to comprehend because K’aila’s mother invokes an epistemology remote from and alien to the epistemology of medical training, scientific research, and formalist Western rationalism in general. The views and the approach of K’aila’s parents probably do seem irrational to the pediatrician, and may seem irrational to many readers of this book, because they proceed from the mysteries of faith and feeling rather than the empirical measurabilities of science and technology. And “true believers” can indeed proceed dogmatically and uncritically, but it is not always unreasonable to involve matters of spirituality in decision making. What is reasonable and what is unreasonable will be multi-factored, context-dependent, and revisable. Whether the decision making of K’aila’s parents was rational was the key issue in the judge’s disposition of the application to have K’aila committed to the care of the Minister of Social Services so that the Minister could authorize and consent to the transplant operation.10 The judge recognized the nature of the choice that had to be made: There is a 100% certainty of death for this child without the surgery. With the surgery there is approximately a 70% to 75% chance of survival in the first year, assuming the child receives a positive assessment and a donor organ. There is between a 60% to 65% chance of five-year survival. Survival necessarily entails an uncertain quality of life. The child will be very susceptible to infection. The child will

40 

Chapter 2

suffer the surgery and the known side effects of the follow-up drug therapy. Death may be avoided and life may be prolonged in the short term. There are no guaranties. (140–141; page numbers in this section refer to the case cited in note 10)

The judge also recognized that medical opinion was divided. On the one hand, liver transplantation is a standard and accepted treatment for endstage liver disease in both adults and children in North America. On the other hand, because of the success rates, the risks of the procedure and of the subsequent drug therapy needed to suppress the immune system, and the impaired quality of life, allowing a patient with end-stage liver disease to die with supportive care and treatment also is a legitimate, medically recognized alternative. The judge then made two important determinations. The first concerned the scope of the issue. The judge recognized that the argument for a transplant is narrowly medical, and he appreciated that a decision about a transplant “cannot be reduced to mere mathematical probabilities” (143). The resistance of K’aila’s parents, the judge also recognized, is grounded on broader practical, emotional, social, and psychological considerations for both K’aila and his parents. Those considerations, in the judge’s view, were decidedly relevant: “It is fair to say that the parents have given a much more profound consideration to the other components in the decision. These other factors must necessarily be given considerable weight” (142). The second determination concerned procedural matters and the nature and quality of the decision making of K’aila’s parents. The judge found K’aila’s mother to be “rational, objective, and balanced in her assessment of the situation” (140). He noted that she “has become thoroughly conversant with the available information outlining the benefits and risks in making a decision” (139), and he pointed out that she had met with two families who have children with transplants and had discussed the situation extensively with them and with various doctors, as well as the director of the transplant program to which K’aila would be referred. The judge concluded that K’aila’s parents “are both intelligent individuals” (139) who after thoughtful and careful consideration have made a fully informed, considered decision to withhold consent for the transplant operation. Given all that, the judge found that their decision making was not unreasonable: The decision reached by the parents cannot be characterized as being overwhelmingly poor, inherently inept, ill-considered, lacking judgment, or simply unreasonable and unacceptable to society. Their decision is not of the quality that the state should be compelled to intervene. … The parents’ choice of care is not unreasonable in these circumstances. (142)

The Difficulties of Applied Ethics 

41

Strikingly, the judge quoted Hippocrates: “Hippocrates said, ‘Life is short, the art is long, timing is exact, experience treacherous, judgment difficult.’ This is just as true today as it was in 400 B.C.” (142). The judge understood that a decision about K’aila’s future required a difficult judgment. Given that understanding, the judge converted the issue from one of substance to one of process: how did K’aila’s parents make their decision, and was their approach, in the circumstances, reasonable? The judge’s inquiry was not purely procedural, however, because he recognized that there are substantive considerations that impose constraints on the decision making of K’aila’s parents. The judge asked, “Can it be said that this is a clear case of rejection by the parents of the values society expects of thoughtful, caring parents for a terminally ill child?” and he answered in the negative. That question raises issues of substance and procedure and itself calls for judgment: for example, which values should be involved here, and when is a case of rejecting society’s values a “clear” case? For present purposes, though, the important point is that the approach it adopts and the constraints it imposes are markedly not the approach and not the constraints of applied ethics. The next section provides another example of how the failings of applied bioethics prompt a shift from substance to process. 2.3  From Substance to Process The manner in which the judge decided K’aila’s fate is distinctive of, but not peculiar to, judicial decision making. Indeterminacies in, and conflicts between, norms—in law, in bioethics, and everywhere else—often make it impossible to straightforwardly derive a decision. Yet a decision has to be made. How does that happen? As in K’aila’s case, the focus shifts from the right decision to making a decision in a right way. John Arras’s analysis of the following vignette illustrates that shift in bioethics. When the norms of bioethics run out, as they quickly and inevitably do, the question of what to do is replaced by the question of how to decide. Vignette 2.4  Mrs. Smith Mrs. Smith, an 85-year-old resident of a nursing home, was transferred to the hospital for treatment of pneumonia. Although she has responded well to antibiotic therapy, her overall condition and prognosis remain grim. For the past three years her mental state has been steadily deteriorating due to a series of strokes, which have finally rendered her severely demented. She is now nonambulatory, incapable of sitting up in bed, and uncommunicative most of the time. When she does talk, her speech is completely incoherent and repetitive. Mrs. Smith shows no signs of recognizing or remembering her family and primary caregivers. The

42 

Chapter 2

nurses in charge of her care assert that she appears to experience pleasure only when her hair is combed or her back rubbed. During her recovery from the pneumonia, Mrs. Smith began to have problems with swallowing food. Following a precipitous decline in her caloric intake, her son and daughter (the only involved family members) consented to the placement of a nasogastric tube. Mrs. Smith continually pulled out the tube, however, and continues to resist efforts to reinsert it. The health care team faces difficult choices regarding Mrs. Smith’s care. Foremost among them is whether her physicians should surgically insert a gastrostomy tube in spite of her aversive behavior. Mrs. Smith has never made a living will, nor has she indicated to family or friends at the nursing home what her preferences would be regarding life-sustaining care in this sort of circumstance. Both her son and daughter have stated that she would nevertheless not have wanted a gastrostomy tube inserted and would, if she could presently decide, prefer an earlier death to being sustained indefinitely in the twilight of her minimally functional condition. In defense of this claim, they note that she has always been a very active, independent person and that she avoided doctors whenever possible.11

The standard approach to deciding what ought to be done on behalf of incompetent patients, in law and in bioethics, invokes two substantive principles: a substituted judgment principle and a best interest principle. These principles never conflict because they are lexically ordered: the substituted judgment principle must be applied first, and the best interest principle is to be used only if the substituted judgment principle cannot produce a decision. So a person deciding about the care of an incompetent patient must ask, first of all, what that patient would want were he or she able to make a decision and to communicate that decision. Often, though, a patient will have expressed no values or beliefs that are applicable to the precise question that needs to be answered, and trying to extrapolate an answer from the patient’s general values or views of life or from the patient’s personality or character traits can yield only speculative, equivocal conclusions. That is the case with respect to Mrs. Smith. Mrs. Smith has not completed an advance directive (or living will) and has not designated anyone as her proxy decision maker. Mrs. Smith’s reported character traits—independence and aloofness from physicians— are compatible with a wide range of interpretations. Nor do her general attitudes, preferences, and tastes provide determinative guidance about the proposed treatment. As Arras notes, “It is one thing to have negative attitudes towards aggressive life support, but it is quite another to actually refuse it in your own case” (1988: 938). Mrs. Smith continually pulls out a nasogastric tube and resists efforts to reinsert it, but how to interpret that behavior also is unclear. It could signal a strong, persistent, and current

The Difficulties of Applied Ethics 

43

desire to reject life-sustaining treatments, or it could be a reflex caused by irritation from the tube. Given these uncertainties, how the substituted judgment principle might apply to Mrs. Smith is merely speculative. Moving to the backup principle, Arras tries to determine what would be in Mrs. Smith’s best interest by using a “moral triangulation” approach that locates her on a continuum of incompetent patients. At one end of the continuum is Mrs. Jones, a patient in a persistent vegetative state (PVS), who has only brain stem activity. She has no awareness of the world or what is happening to her and is unable to feel pleasure or pain. Mrs. Jones’s only interest seems to be the remote possibility of misdiagnosis. Because Mrs. Jones’s lack of conscious experience renders her incapable of being harmed, she cannot be said to be excessively burdened by treatments or to be “better off dead.” So a decision to cease treating Mrs. Jones would have to be based on the conclusion that she no longer is a “person” (Arras 1988: 940). Patients at the other end of the continuum do have interests, however. Arras’s example is Mr. Black, a 90-year-old marginally functional patient who is suspected to have colon cancer but who refuses a laparotomy to confirm the diagnosis. Mr. Black seems to be competent, but psychiatric examinations disclose profound deficits in his short-term memory and substantial confusion about his condition and his surroundings. He is described as “pleasantly demented.” Because Mr. Black clearly is a “person” with interests that can be promoted or frustrated, decisions about his treatment would be based on a patient-centered best interest test (Arras 1988: 940). Where does Mrs. Smith fall on this continuum? She is somewhere between Mrs. Jones and Mr. Black. Like Mrs. Jones, Mrs. Smith no longer has an enduring “self” and has been reduced to “a mere locus of transient sensations.” Like Mr. Black, though, she possesses some conscious experience in her ability to feel pleasure and pain and to have perhaps rudimentary emotions. Given the similarities to Mr. Black, one could try to ascertain what is in Mrs. Smith’s best interest. To justify termination of treatment according to the best interest test, two criteria have to be satisfied: (1) The burdens of a patient’s life with the proposed treatment would clearly have to outweigh whatever benefits might be derived from continued life; and (2) Further treatment would have to be inhumane due to the presence of severe, uncontrollable pain. With respect to the first criterion, it is hard, as Arras points out, to discern any burdens of continued treatment for Mrs. Smith: “In the absence of any persistent and severe pain … it would appear highly doubtful that the burdens of Mrs. Smith’s continued existence clearly and markedly outweigh the benefits, even when the benefits

44 

Chapter 2

approach zero” (Arras 1988: 942). So the decision that follows from the best interest test is that the G-tube should be surgically implanted in Mrs. Smith and her life should be maintained indefinitely with artificial nutrition and hydration. That result seems decidedly unpalatable because the best interest test focuses too narrowly on pain and ignores other important moral considerations such as privacy, dependency, dignity, and bodily integrity. Unfortunately, Mrs. Smith has no awareness of how her dignity might be despoiled or her privacy violated. Any appeal to such concerns would have to invoke Mrs. Smith’s prior preferences by claiming that she would not have wanted to be so utterly and completely dependent; that she would not have wanted to have her privacy violated so intrusively; and that she would not have wanted to have her dignity so deeply insulted. But that argument represents a return to the substituted judgment test, for it is an attempt to figure out what Mrs. Smith would want were she presently able to decide for herself. The attempt to apply the principles of substituted judgment and best interests is going in a circle: the inconclusiveness of the substituted judgment test leads to the best interest test, but dissatisfaction with its narrow focus leads back to the substituted judgment test. How does one escape that circle? Arras’s proposal is to shift from substance to procedure. Usually, in situations involving incompetent patients neither the substituted judgment test nor the best interest test can be applied given the sparse, tenuous, and ambiguous evidence that is available and the diverse conclusions that caring, sensitive, and responsible people are likely to derive from that evidence. So the question that must be asked, Arras contends, is not what should be done, but who should decide, and that determination is central to the decision-making process. Arras concedes that there are clear cases of overtreatment and undertreatment of patients but points out that between these extremes there is “a vast gray area …where the patient’s best interests will remain unclear and largely inscrutable” (Arras 1988: 943). In this morally nebulous realm, the key moral question becomes who should decide about the care of the patient, and Arras’s answer is that knowledgeable, concerned, well-intentioned family members are in the best position to decide what is best for the patient. Arras adds two qualifications, one of which is uncontroversial and one not. The uncontroversial qualification is that, as in the case of K’aila’s parents, there is a substantive constraint on this decision making: “the decision of a trustworthy surrogate should prevail over objections from caregivers, unless the latter can show a clear violation of best interests.” The controversial qualification concerns what kinds of considerations are

The Difficulties of Applied Ethics 

45

morally relevant to surrogate decision making. Arras acknowledges that “many caring and well-meaning family members will also want to weigh the impact of continued treatment upon themselves and the family unit,” and he endorses that position: “This is to be expected and should not give us grounds for concern so long as the case originally falls within the gray area of ethical ambiguity, and so long as the interests of family do not clearly violate the best interests of the patient” (Arras 1988: 943; emphasis in original). Shifting from the application of substantive norms—what to decide— to the design of decision-making processes—how to decide—is not just practically helpful, however. It also is philosophically attractive because, as we will see, it protects bioethics from the arbitrariness and subjectivity that ensues when principles, whether substantive or procedural, cannot be applied to produce definitively correct answers. Arras’s procedural norms suffer from the same indeterminacy as the substantive norms of applied ethics. For example, they cannot sharply circumscribe the boundaries of that capacious “gray area of ethical ambiguity”; cannot comprehensively identify which members of families are “involved” and “well intentioned”; and cannot precisely instruct families about when and how to consider their own interests as they make decisions about the care of an incompetent relative. As with substantive matters, procedural decision making often requires judgment. The deeper lesson of Arras’s analysis is that, given the inevitability and prevalence of judgment, which is required whenever there is no simple answer, the rational alternative to inaction or arbitrary action is to design a right way to make judgments. That is what the women making decisions about whether to try to become pregnant after receiving genetic counseling did, that is what K’aila’s mother did, and that is what the judge endorsed. That is also what is done outside of ethics, not only in law but everywhere that judgment is similarly required: in educating the young, in designing a new building or ornament, in solving crimes, and so on. In taking this approach we ground our ethics in a universal form of intelligent problem solving adequate to the tasks (see chapter 4). 2.4  The Scope of Morality and the Limits of Formal Bioethics The principles of bioethics are to be applied to moral problems, but what makes a problem a moral problem, moreover, one to which the principles of bioethics are pertinent? Does the following vignette pose a moral problem, and if so, are the principles of bioethics relevant to it?

46 

Chapter 2

Vignette 2.5  Dr. P Dr. P, who works as an internist in an inner-city neighborhood, is the physician for the 12-year-old grandson and the 9-year-old granddaughter of Ms. J. Ms. J’s son is a homeless heroin addict. The grandchildren’s mother, who was also a heroin addict, is dead. Because Ms. J, who is 79 years old, is becoming progressively more demented, it is not clear how much longer she can care for her grandson and granddaughter. Her son is also Dr. P’s patient, but she is not. Ms. J participates in the Grandparents Who Care program at Dr. P’s public health center, and Dr. P facilitated a support group to which Ms. J belonged. Dr. P also has done respite care for the family and has taken the children on several outings. Even though Ms. J is not officially her patient, Dr. P cares about her, feels an obligation toward her, and “treats” her in the context of a grandparent support group that Dr. P started because so many of her patients need psychological and material support as well as medical care. Dr. P agonizes about the boundaries of her medical and personal responsibility to Ms. J and her grandchildren, and she has even entertained the possibility of adopting the two grandchildren (Kaufman 1997: 11).

Many people would regard Dr. P’s problems—or perhaps Dr. P’s perception of her problems—as personal or psychological or emotional, but not moral. Yet the vignette poses fundamental medical-moral questions about who is a patient and what should be treated. These questions would not be recognized as moral questions relevant to medicine by those who see medicine as a separate scientific and professional enterprise; rather, Dr. P’s worries would be taken simply to expose her misconceptions of the basic nature and goals of medicine. Nor would they be recognized as moral questions relevant to bioethics by those who equate bioethics to the application of principles. Moreover, Dr. P will be aware that she could not extend such care to every similarly needy family she might treat and that her self-assumed responsibilities are specific to her history of engagement with this particular family, and that context dependence would be taken to show that no moral issues are involved here because the norms and decisions of morality must be universalizable. But the questions Dr. P finds agonizing challenge the assumption that medicine and bioethics can be cleanly demarcated and segregated from particular, personal practice. They are questions about what medicine is, what a commitment to being a doctor means, and what the limits of a physician’s responsibilities are. They are questions that blur the boundaries between the medical, the social, and the moral—questions in which clinical matters, family matters, and moral matters are intimately and inseparably entangled. There is this to be said for recognizing Dr. P’s struggle as a moral struggle. Dr. P is troubled in the same way that the women who received genetic

The Difficulties of Applied Ethics 

47

counseling are troubled: she is trying to figure out how to live well. She is a physician whose professional role is central to her identity and her integrity as a person. She is questioning her role as a physician because being a good physician is, for her, crucial to being a good person. Applied bioethics does not recognize her struggle as a moral problem because it does not conceive morality in terms of the good life, living well, and flourishing in particular circumstances. Applied ethics is a problemsolving tool and, as with any tool, it works only where the conditions are suited to its application. That requirement specifies and constrains the domain of bioethics problems. But why should such a truncated view of bioethics—and morality, more generally—be accepted, and does such a view capture why moral problems are so troubling and so momentous? 2.5  Conclusion The vignettes in this chapter convey two crucial lessons about the nature of morality. One is that there is much more to morality than principles and rules. Principles and rules are an important component of morality, but far from all of morality. The other is that judgment is pervasive and inescapable in ethical decision making. Most of the time principles and rules, in concert with values, customs, and practices—all embedded in social, political, economic, religious, and legal contexts and in background worldviews—guide our behavior with little or no thought; real morality is like the air we breathe. When moral problems arise, however, principles and rules, isolated from the complex, multiple contexts within which they are understood and operate, and with all their weaknesses, are of little help. The morally relevant description of a problem has to be constructed, the salience of principles and rules established, the scope of general terms and concepts determined, and conflicts between principles and rules resolved. Judgment is required for all of that, and once those judgments are made, the appliedethics exercise of subsuming facts under norms is but a logical flourish. What the deliberations of the women after genetic counseling in chapter 1 reveal, and what the agonizing of K’aila’s mother and Dr. P in this chapter reveal, is that the abstract norms of applied ethics miss what real moral problems are and what they mean to real people. Real moral problems force people to make choices about how to live well—and to die well—and about what kind of person to become and to be. People take those choices seriously, even if they do not consciously regard them as moral choices, and they deliberate about them seriously, albeit not in the manner prescribed by formal applied bioethics.

48 

Chapter 2

The core question for morality, then, is: how can people rationally deliberate about moral issues and make moral judgments about them? Applying principles or rules to deduce universally correct answers, as happens in logic and mathematics, our paradigms of rationality, is the hallmark of rationality in the Western analytic philosophical tradition. Because a main goal of moral philosophy is to establish the rationality of morality (or else threaten its death in arbitrariness), it naturally aspires to, and thus tries to emulate, formal reason. The consequence is that morality becomes applied ethics. But the endemic problems of applied ethics elude that kind of rationality—and in similar ways, ironically but instructively, as does the invention of logical and mathematical systems themselves. The attempt to avoid these problems by combining principles and rules into a moral theory that resembles mathematical theories also fails because judgment is inescapable in that endeavor too. Whence the philosophical enterprise of using formal reason to vindicate the rationality of ethics fails. That outcome leaves two options. One is to maintain the equation of rationality with formal reason and concede the arbitrariness and subjectivity of the judgment that pervades not just morality but our lives. And the other alternative? We know that cognitively we are capable of dealing with moral problems, replete with all their entangled issues, context dependence, conflicts, and shades of meaning, because we see it done, in the women responding to genetic counseling and in ourselves from what we accomplish throughout our daily lives, for example, as we constructively resolve a dispute between our children, or juggle parental and work roles. The alternative then is to develop a more expansive conception of reason, along with its corresponding epistemology, that matches the basic cognitive processes involved and that can thus account for the rationality of judgment, in morality and in our lives. This latter, preferable option, and the ethics that accompanies it, is the project of this book.

3  The Problems Generalized, Diagnosed, and Reframed in a New Account of Reason

C

The widespread presumption in ethics is that moral reasoning is a matter of applying universal moral principles to the relevant descriptions of the particulars of situations and thereby deducing right decisions about what to do. The preceding chapters have presented the many practical and theoretical inadequacies in this depiction of ethics. What should the response to these serious defects be? Simply rejecting the model of applied ethics will not solve the problems. Doing that would leave the sources of the difficulties undiagnosed and provide no constructive alternative. Anticipating, we believe that the twin errors from which the grip of applied ethics derives are mistaken conceptions of the fundamental natures of reason and of morality, and that we must dig deep to expose those root errors and use them to construct better alternatives. However, the abstractness of the issues, combined with the seeming obviousness of the conventional wisdom, make this task difficult. We begin by briefly examining the parallel situation in science, namely, the presumption that applying science to a situation is a matter of conjoining universal scientific principles to the relevant descriptions of the particulars of situations and deducing predictions and explanations from them. Our aim is to show how the same practical and theoretical inadequacies that exist in applied ethics are also inherent to this model of application in science. But in science their deeper origins are more clearly visible. It is generally believed that ethics and science have little or nothing to do with one another because they pursue quite different goals—knowledge, in science, and goodness (or rightness or virtue) in ethics—and they employ quite different methods in doing so—logical principles and empirical observation are used in science; moral principles and normative intuitions in ethics. To the contrary, though, once the same root inadequacies have been exposed in both science and ethics, we can be confident that their source is a shared but inadequate conception of rationality, and once that

T

50 

Chapter 3

conception is replaced by a more adequate one, the similarities between science and ethics will become clearer. 3.1  The Parallel Difficulties of Applied Science For convenient comparison, scientific analogues of moral problems presented in the preceding chapters are given, followed by brief ethical parallels. 1. Problem: the indeterminacy of what is relevant in a situation. Science: the indeterminacy of what methods, principles, and models are predictively/ explanatorily relevant for a particular situation. In the case of research into the behavior of plasmas (very hot ionized gases), for instance, it is not initially clear whether standard statistical analyses are valid—perhaps free electrons are not randomized and so do not properly have a temperature— or whether wire probes (the crude but most usable method then available) with their interposing distorting sheaths would permit their sheath properties to be distinguished from plasma properties, and so on. In manifold similar ways each distinct situation raises anew the issue of what experimental methods, principles, approximations, boundary and initial conditions, and so on are relevant for understanding it. Ethics: each situation raises anew the issues of what features of the situation are morally relevant and what principles, values, situational characterizations, and so on are required to properly understand it ethically and so validly determine what ought to be done. In each case these issues cannot be settled by formal considerations alone and are instead matters for sound nonformal judgment. 2. Problem: the initial indeterminacy of which specific features of relevant factors best characterize a situation. Science: the unfolding understanding of the predictive/ explanatory issues as the situation is investigated. Initially, for instance, plasmas were simply probed by single wires, but later multiple probes were developed that provided improved information, in turn allowing plasma temperatures and containment geometry to be chosen to obtain the clearest measurement conditions, and so on. This process of interlocking improvements is typical and central. Ethics: each new situation raises the issues of what is the most appropriate characterization of the problem, what methods should be used, and what kinds of solution would be acceptable. Coordinated understandings of each of these dimensions of an ethics problem unfolds as the situation is investigated.

The Problems Generalized, Diagnosed, and Reframed 

51

Again, formal notions of reason like deduction and induction cannot, in principle, offer guidance. 3. Problem: the incompatibility of principles. Science: the incompatibility within and between theoretical and methodological principles. In the plasma case, for example, one can model the hot interior as either a fluid or a gas, and may model it thermodynamically or dynamically, but not two or more combinations of these at once. Wise scientists select now one, now another as appropriate to the current situation. Ethics: the incompatibility within and between theoretical and methodological principles. Deontological principles, for example, “do no harm” and “tell the truth,” conflict with one another. Methodologically, deontological rule-following and utilitarian utility-maximizing produce conflicting outcomes, for example, between freedom and welfare over issues like vaccination, voting, and vagrancy. When relevant premises for reasoning conflict, any deduction or induction can lead to arbitrary results and vitiation. What is required is judicious contextual selection of fewer consistent premises or modification of premises to construct a compromise among them (see chapter 6), but formal notions of reason cannot, in principle, guide that process. 4. Problem: the incompatibility among basic values. Science: the problem of incompatibility among basic epistemic values. In plasma research, there are situations where the dominant value is the refinement of probe empirical accuracy and others where the dominant value is valid testing of theoretical explanation (e.g., testing for the predicted occurrence of electron cyclotron resonance), but optimizing for each of these values requires distinct setups, modeling, procedures, and data treatments. Such conflicts can arise between any pair of the many epistemic values of science (e.g., among any pair of empirical adequacy, explanatory power, theoretical unification, or technological applicability). Ethics: the problem of incompatibility among basic ethical values. Ethically salient actions aim to promote values such as autonomy, integrity, equality, justice, welfare, happiness, pain-freeness, and esteem.1 But attempting to pursue these values typically results in multiple conflicts (notoriously, increases in total welfare often come at the expense of decreases in equality, and vice versa). This is another version of problem 3 and the same conclusion drawn in 3 above also applies here. 5. Problem: the occurrence of deep, ill-defined problems. Deep problems are those for which not only is the solution unknown, but the problem itself

52 

Chapter 3

is not well defined, so that the values that ought to drive its investigation and the valid methods to do so are unknown, unclear, or in dispute, as are the set of applicable theoretical models, the solution set, and the criteria for successful resolution. Science: deep problems abound wherever a sufficiently novel research domain raises anew the applicability of extant concepts, models, and methods. Astronomical investigation with Galileo’s telescope was a deep problem because uncertainty surrounded how telescopes worked, what they could reveal, and so on (Feyerabend 1978). The same is true of plasma investigation with probes. Conversely, the emergence of particulars of data, theory, or method can transform conceptions of what the problem and the solution goal are, as did Galileo’s observance of the moons of Jupiter and their phases. Ethics: deep problems abound in social life, for instance, ranging from how to throw a nice party for new immigrants, to how to regulate cloning.2 And our conceptions of such problems can be transformed by the emergence of particulars of description, theory, or method, as when we discover that some deviant behavior by a teenager is motivated by boredom and not rebellion. Formal inference is defeated by ill-definedness, and judicious nonformal judgments are required both to develop the missing information, infrastructure, and so on, and to understand their significance. 6. Problem: context/situation-dependence and universality. Science consists of concepts, theories, infrastructure, and methods that range in generality from universal (e.g., laws of dynamics) to highly specialized (e.g., genetically clean artificial insemination of monotreme ova) and everywhere between (e.g., Galileo’s and electron microscopes and organic chemistry of the intracellular Krebs cycle). The more specialized support the more general while the more general guide the construction and validation of the more specialized. For example, an organic chemistry model of the intracellular Krebs cycle will employ some general chemical techniques such as determining assays, while also employing highly specific and specialized concepts and techniques (i.e., methods) concerned with intracellular chemical dynamics; it will rely both on general background knowledge drawn from physics, chemistry, engineering, statistics, and on methodology specialized to that context.3 All claims are accepted into the body of scientific knowledge, but each on grounds appropriate to its class of valid contexts. Appropriately, scientists in flourishing sciences maintain many highly context-dependent specialized relationships with other scientists, often distant in expertise. Wise scientists aim for (feasibly achievable) generality but select and specialize activities, arguments, and relationships to suit their total context.

The Problems Generalized, Diagnosed, and Reframed 

53

Ethics consists of concepts, theories, infrastructure, and methods that range in generality from (presumptively) universal (e.g., principles of justice and utilitarianism) to highly specialized (e.g., a woman deciding whether to try to get pregnant after having received genetic counseling) and everywhere between (e.g., third-party privacy walls for acquisition of sensitive medical information). The more specialized support the more general, while the more general guide the construction and validation of the more specialized. Flourishing is characterized and guided by the relationships, practices, and structures within families, communities, countries, societies, and cultures. Appropriately, those who pursue ethical understanding maintain many particular and highly context-dependent relationships with other ethicists, legal experts and scientists, often distant in expertise, as well as lay persons of diverse experience. Wise ethicists aim for (feasibly achievable) generality but select and specialize activities, arguments, and relationships to suit their total context. Because rule-bound rationality ignores the profound dependence of flourishing on both situation and context—universal logical principles are necessarily independent of context—it is unable to contribute much to the process of flourishing. Drawing together the many hints about the proper terms in which to discuss the natures of science and ethics that have emerged during the foregoing discussion, we briefly summarize the deep similarities between science and ethics, under four headings: problem solving, judgment, normative learning, and historical development. Problem solving. All these parallels suggest that the conventional wisdom about the disparity of science and ethics needs to be re-thought. Both engage in normative pursuits to determine what ends are relevant and what ought to be done to pursue those ends. So they share the same overall procedural, problem-solving framework. Using this framework in either discipline involves the capacities of being able to recognize relevant problems and what is normatively important about them; recognize and assess methods for exploring opportunities for resolution; assess the outcomes; iteratively modify problem, method, and resolution accordingly; and the like, and do all that in context-sensitive ways that ensure the flourishing of both the normative endeavors embedded in, and the particular persons involved in, the institutional arrangements. Judgment. Working within that common problem-solving framework and addressing the six general problems that science and ethics share (see above) requires making a crucial set of nonformal, skilled judgments of

54 

Chapter 3

the kinds indicated in the preceding paragraph (expanded list is provided in section 4.1). Any adequate conception of reason has to encompass the rich variety of reasoning processes that we in fact use to intelligently recognize, characterize, and resolve our problems—from which it follows that we require a sufficiently rich set of reasoning skills, both substantive and procedural, to execute the foregoing tasks. It seems clear that those processes cannot be grounded in formal reason.4 Judgment is a core human skill that is largely irreplaceable by rational formalisms. Judgment is rationally learnable and improvable like other skills, and, like all intelligent creatures, humans need abundant practical experience with contextual problem solving for judgments to be sufficiently relevant, attentive, perceptive, and judicious to support intelligent problem solving. This requirement clearly holds for individuals and is reflected in the intensity and length of practical training for any high-level skilled role, obvious in medicine and law where preparation involves both academic work and supervised experience; it should equally be reflected in training for all ethical decision making5 as well as for research and for military or public service roles, in apprenticeships for skilled trades, and so on. Normative learning. Equally telling for the role of judgment, even in learning about our norms, is that norms and normative methods have been massively elaborated over time in both science and ethics as experience has been gained, in a process characteristic of knowledge and skills that the human species acquires through learning. In science, Galileo, Newton, Bacon, and many others had to develop, systematize, and scrutinize elementary methods for obtaining empirical knowledge, even after 2,000 years of discussion of the idea. The historical fruits of these and subsequent labors are manifest in today’s situation, where statistical inference, decision theory, and the like provide us with powerful, proliferated, and refined methodological tools. And norms governing acceptable procedures and outcomes/solutions have been greatly extended and elaborated: for example, in science into the 24+ array of epistemic values now recognized (see note 22). Nevertheless, no method has been found to apply universally unconstrained by other norms and methodological challenges remain as science develops, provoked, for example, by the deep and widespread consequences of the recent appearance of complex systems models across virtually all the sciences.6 Similarly, contemporary bioethics has emerged from the developing practices of medicine and biology and has been spurred on by the aberrations of wartime research on human subjects and by profound social movements directed at eliminating discrimination and protecting rights. Simple principles and codes of ethics have been elaborated into

The Problems Generalized, Diagnosed, and Reframed 

55

complex amalgams of standards, guidelines, and policies, legal precedents and regulations, and case analyses—a process supported by university training programs, research institutes, and professional societies, conferences, and journals, just as it is in the institutional development of science. As for ethics, then, consider how bioethics has recently undergone an important shift, from the narrowly focused ethics of doctor-patient relations considered in isolation, to an explosion of institutional ethics committees with wide memberships and broad institutional domains and the emergence of public-health ethics for policy formation and social interventions (e.g., for management of epidemics, vaccination programs, and the like). Of course, many problems will still be generated from the doctor-patient locus as well as from other institutional loci such as government-sponsored health programs. Correspondingly there has been an expansion of formal professional ethical and spiritual advisory roles to private firms as well as public institutions. Conversely, these institutional frameworks both facilitate and shape normative learning in the individuals who develop within them.7 These are all invented social processes, creative constructions which we are learning about (as well as learning from) and elaborating as we go. Historical development. Finally, the histories of science and ethics reveal key shared evaluative or normative features that pertain to an understanding of their natures. For each domain, several competing fundamental theories about its normative structure have emerged; for example, empiricist confirmation and Popperian refutation in science, and foundationalism and coherentism in ethics.8 Yet each of the underlying simple models of reason and knowledge—for example, empiricism in science and rationalist foundationalism in ethics—fails to provide an adequate foundation. Consequently, alternatives have proliferated that, in various ways, reject being confined to narrow conceptions of reason and knowledge—for example, Feyerabendian and social constructivist accounts in science and casuistry and narrative accounts in ethics. In each domain a complexly interrelated set of irreducible values has emerged that guides the pursuit of problem resolutions, for example, empirical accuracy versus generalizability in science and equality versus need in ethics, the values tend to conflict with and exclude one another in practical situations. In short, in neither domain is there a single, satisfying normative conception of formal problem solving, let alone one that simply prescribes the application of principles or rules to problems to derive solutions. Yet in both science and ethics just this is what standard practice presumes. The conditions for the resolution of deep problems typically are not given in advance but are discovered or constructed through the process

56 

Chapter 3

of trying to resolve those problems. Experienced practitioners in science and ethics (and elsewhere) have learned how to do this kind of problem solving effectively. Their success is commonly attributed to good luck, serendipity, inexplicable revolutionary illumination, or mastery of an art. In fact, it is attributed to anything but a rational process of problem solving. Why? Because rationality is presumed to be restricted to formal rulegoverned formulae for making decisions. But that narrow conception of rationality simply evinces the prejudice toward formal reason. To the contrary, the fundamentalness and ubiquity of deep problems, combined with widespread success in solving them, makes a strong case for inverting that bias and presuming instead that solving deep problems represents the pinnacle of our rational intelligence. Given that realization, the challenge is to develop a more expansive conception of reason that can support such a capacity and its achievements. The reward would be an account of reason and a correlative problem-solving process that would span the disparate domains of science and ethics, and hence have a claim to fundamentalness. It would also explain our proper conviction that it is more intelligent to invent useful rule systems and demonstrate their domains of validity and incompetence than merely to use them. Most important, it would open up much richer conceptions of problem solving in science and ethics that are better suited to the multidimensional, integrated problems that characterize our more complex societies. We present our first steps toward such a model and its philosophical framework in chapter 4. To close this discussion we briefly consider the option of comparing the difficulties of formal applied ethics with those, not of science, but of social and political sciences, cultural theory, and anthropology. These more personal-social domains all seem closer to applied ethics as social practice than does scientific research, with its focus on empirical knowledge rather than social behavior and its regimented theoretical and experimental structures pursuant to that end. We agree that the more directly social domains are of especial concern to ethics, and that there is more sympathy to be found in the more social domains than is likely to be found in traditional philosophy of natural science for the kind of nonformal approach to reason that we pursue here. However, we offer three primary reasons for instead choosing science as our reference point. (1) Assumptions concerning the nature of reason run deep; often they are so deeply embedded as to be largely or wholly tacit, their users unaware of adopting them. In particular, assumptions about the role of logic are often considered so obvious as to require no defense. In this respect, the lack of contrast between ethics and the various social action domains, given their shared focus on persons and

The Problems Generalized, Diagnosed, and Reframed 

57

social action, would tend to obscure the underlying assumptions. That science is so evidently different from ethics makes it easier to compare the uses of reason in the two disciplines. (2) We have a much more explicit, precise, and detailed account of reason’s roles in science than in any of the personal-social sciences. This makes it possible to more readily discern the similar account that must also underlie ethics. For instance, the formalist ethics paradigm of deduction of right action from principles and contextual information (in ethics) comes into sharper focus once we see its comparison with prediction and explanation (in science) as deduction from theoretical principles and contextual information. And in particular, we have a much more explicit, precise, and detailed account of formal reason’s failures in the realm of science than we have in any of the personal-social sciences, making it easier to characterize a richer nonformal conception of reason and its applications. (3) The personal-social domains present us with large internal diversities and less of a tradition of deep, integrated development of positions with which to mount a proper comparison. This makes addressing the nature of reason in these domains a long, tortuous task, and one that has less cognitive force. For all these reasons, proceeding as we do is more practicable, forceful, and illuminating. Moreover, one of us (CH) has developed the view that all domains of problem solving share the same core cognitive process (see Hooker 2017a, 2017b, and briefly the cases in section 5.2). If that were so, it would undergird the comparison with science and provide a common framework for developing the correlative processes of nonformal reason and of deliberation and judgment formation. 3.2  Confronting the Dominant Analytic Tradition The preceding discussion has been about setting up problems properly—a necessary precursor to solving them—and that discussion provides powerful motivation to pursue an alternative, more adequate conception of reason than its equation with logic or formal inference. But two substantial considerations militate against this endeavor: the seeming obviousness of logic as central to reason, and, equally, the seemingly self-evident nature of the Western conception of rationality (see below) that frames that preference. The former deems the project of seeking alternatives improper; the latter renders attractive alternatives wrongheaded, if not unthinkable. So, to complete our preparation for thinking afresh about reason, this section briefly summarizes the limitations of viewing formal calculi, and logic in particular, as the primary expression of reason and then considers the related limitations of the Western conception of rationality that underlies

58 

Chapter 3

the focus on logic. The analysis is brief because the aim is not to provide an exhaustive critique of Western thought about rationality, but simply to convey the essence of its difficulties so that our positive development of a more expansive conception of reason can proceed. At the same time, though, the ensuing critique of formal reason provides valuable guidance for the construction of post-formal (i.e., nonformal) reason. 3.2.1  The Critique of Logic as Method Formal logical reason assumes that rational procedure is confined to the use of valid formal arguments—i.e., arguments whose validity is determined by universal logical rules—where the use of those arguments to construct inferences is compelling (Descartes’s clear and distinct ideas), not least because their normative power derives from their form alone. In deductive logic all and only valid arguments are those for which the equivalent argument form is a tautology, a proposition that is true by virtue of its form alone.9 This strictly separates rationality as form from any content, elegantly entailing that reason cannot introduce any content errors of its own into the exercise of rational deliberation and decision making. That guarantee has been the irresistible attraction of formal reason for the certainty-oriented Western philosophical tradition.10 Logically valid arguments are indisputably part of rational procedure, and any account of rationality that cannot appropriately accommodate them is defective. The crucial issue, however, is whether logically valid arguments are the whole of rational procedure, or even indicative of the whole of rational procedure. It has seemed obviously so to many, so obvious that Boole described logic as the “laws of thought.” Others have thought that the necessity which the premises of a valid argument convey to its conclusion is so strong that it must derive from a transcendental warrant rather than, as in science, from the support of successful practice. Moreover, since deductive logic is necessarily truth preserving, and truth is the chief goal of rational intelligence, logic is taken to be the rational-making feature of intelligence. The dominant logic-based philosophical tradition in both science and ethics has taken as its goal to show how each domain is rational because it conforms to logical reasoning. The idea that humans are rational animals because they are animals (additionally) equipped to reason logically and thus to pursue truth and practical advantage, has been part of the core Western tradition since Plato and Aristotle. It has been recently rejuvenated with a new plausibility and suggestion of universality by the metaphor of intelligence as computation. Practical rationality is limited, however, because the ends humans choose to pursue have to be declared

The Problems Generalized, Diagnosed, and Reframed 

59

nonrationally determined; but thereafter the means they choose to pursue those ends can be rationally determined according to their formal efficiency, for example, by cost-benefit maximization. Yet despite its elegance and power, the conception of reason as formal just as obviously does not stand up to examination. First, because the invention of logics would not itself count as rational, removing rationality as central to intelligence. Second, because of all the roles it cannot fulfill which require nonformal judgment instead, noted in section 3.1 above. And third, because neither of the twin processes at the heart of scientific knowledge formation—induction and refutation—is fully capturable in formal logic. Consider first scientific induction. The inductive process of arriving at a confirmed theory T on the basis of evidence E is well known to be equivalent to making a deductively invalid, hence illogical, argument: if T (a theory) is true, then, if initial conditions I obtain (e.g., those specifying an experimental setup), then empirical evidence E is true, and I and E are true, so T is true, which has the invalid deductive form “T entails (I implies E), and I and E, so T.” Attempts to construct formal inductive logics to legitimize scientific inductive practice have been made, but despite the development of various formal calculi, three basic obstacles to such a theory of induction endure. (I) Formal form-alone inference cannot introduce any concepts not found in the premises, so any inductive inference to theory from observation will be excluded because theoretical terms are all unobservable so cannot appear among the data. (II) Inductive inferences cannot always be consistently conjoined to other inferences, deductive or inductive. Whence induction is irretrievably local, it has no capacity to generalize.11 (III) Reasonable inductive practice does not support the formcontent distinction required to form a logic because reasonable inductive form is a function of empirical knowledge.12 So the first central intelligent process within science falls outside of formal logic. Now consider refutation, the opposite experimental situation in which the evidence contradicts the theoretical prediction. The simplest logical expression of refutation is deductively valid: “T entails (I implies E), and not E, so not I or not T.” The epistemic point of refutation, however, is not to record the mere appearance of a logical inconsistency and hence of an error somewhere, but to localize and thereby identify errors in order to eliminate them and thus improve knowledge. It is normally assumed that I and E are known because observed, but observation is notoriously fallible so a possible site of error. Logical inconsistency, the only resource available here to formal reasoning, cannot locate an error, let alone identify its source. So

60 

Chapter 3

far as inconsistency goes, error lies in either I or T or E, or in any pair or in all three, i.e., anywhere is possible. Each option can perhaps be checked further, but if this search is done by uncovering inconsistencies, error localization evaporates into an endlessly proliferating cloud of possibilities. In sum, error identification and localization also must be reckoned an intelligent nonlogical procedure.13 Combining both considerations, logic cannot model revolutionary scientific transitions, such as that from Ptolemaic to Newtonian dynamics, because they creatively overthrow (cf. refutation) basic concepts and principles and replace them with improved ones (cf. confirmation), and that cannot be a purely logical process for all the reasons set out above. Revolutions are the resolutions of deep problems, and logic cannot resolve such problems. Notoriously, revolutions have ultimately been considered irrational for all these reasons, despite constituting our greatest cognitive achievements in science. That inappropriate outcome we can, and will, change (see chapter 4). More difficulties could be added, but here we conclude by noting that science is most powerfully formulated mathematically, and mathematics has a history of (re)construction every bit as thoroughgoing and counterintuitive as that of science itself,14 one that includes the development of all the different logics of recent times (modal, fuzzy, relevance, multivalued, quantum, …). How is that history rational? It cannot be a logical induction from experience without revisiting those problems, and it is no more plausibly a sequence of a posteriori recognitions of a priori truths than is science. In fact logic appears to have the same kind of fallible construction dynamics as science has. Contradicting this, since Plato, philosophy has been tempted to consider mathematics, including logic, a priori, an intuition of the necessary, part of a normative/factual dichotomy that runs across every domain and is intended to keep the normative uninfected by the empirical that Plato so distrusted. But in fact there has always been a strong interaction between normative and descriptive theories, which, both theoretically and practically, the sciences have forced and which that dichotomy makes incomprehensible.15 It has not proven possible to defend this divide in any area. In particular, it has not proven possible to isolate rationality theory from developing science because ultimately our theory of our rational capacities must be shown to be grounded in the actual operations of our embodied minds, to be possible developmentally through our actual educational/scientific practices, and to be biologically feasible. This interaction of the putatively transcendental with the factual invites a more naturalist approach in

The Problems Generalized, Diagnosed, and Reframed 

61

which theories of both normative and empirical subject matters are treated as fallible and as being learned and revised in essentially the same manner, and thus as capable of interactively developing in the ways we see them actually doing. In short: Any adequate conception of normative knowledge has to be able to encompass the rich variety of normative judgments and learning processes that we in fact use to intelligently frame and resolve our problems and thereby to improve our knowledge. When logic is thus viewed as a theory of reason to be judged against our actual best practices, it thoroughly fails to satisfy this principle. This quick survey implicates the restriction of rationality to formal logic as the source of the difficulties with rational procedure in science and ethics. Recognizing the parallel difficulties in applied reasoning across two large areas of human life, science and ethics, should instill confidence in the broader generalization that the same difficulties occur wherever we attempt to apply general knowledge. A trial of situations in areas other than science and ethics will quickly confirm that universality for the reader. Conversely, the universality of the difficulties suggests that their root lies deeper than the particular natures of their domains, and that we should look for it not in the peculiar complexities of scientific, ethical, technical, or other forms of decision making but in the constrained, formal nature of the reasoning that underlies the rule-bound, deductive conception of applied knowledge.16 For many thinkers, though, it is difficult to even conceive of an alternative to rule-bound rationality: the assumptions of the dominant Western philosophical tradition about the nature of rationality are so deeply ingrained that they sometimes are invisible. Deconstructing that tradition and showing how there can be alternatives to it should help us to see beyond those assumptions. 3.2.2  Critique of the Western Philosophical Tradition In antiquity the rise of philosophy reflected a distinctive opposition of reason to custom, religious commitment, and brute power, a commitment to a life founded on knowledge and practical expertise developed with and grounded in reason. For our purposes here the essential ideas of the Western intellectual culture that emerged are twofold: (1) humans are rational animals, and it is their reason that distinguishes them from other animals; and (2) the essential project of reason is transcendence of imperfection— ignorance, prejudice, bias, egocentrism, anthropocentrism, anthropomorphism, and the like. Imperfections are transcended through a process of

62 

Chapter 3

reasoning that improves our state or condition with respect to our ideals (cognitive, moral, and aesthetic ideals, and perhaps others). Transcendence typically is taken to be effected through contact with an external reality that involuntarily conveys true content. The involuntariness indicates that we are not fooling ourselves with anthropocentric and egocentric projections or wish fulfillment, distorting the truth with bias and prejudice, and so on.17 For rationalists, contact is with some transcendent world of abstract truths or forms that delivers certain knowledge of the constitutive norms characterizing a domain and the substantive content in that domain. In ethics, for example, substantive ethics is conceived as a system of nonempirical principles and rules of conduct, or a set of virtues. For science, substantive knowledge is empirical and contingent, deriving from the involuntary outcomes of contact with the external material or internal phenomenal world. Metaethically, normative domains could be comprised of either rationalist necessary knowledge or nonrational collective conventionalism and/or individual subjectivism.18 Most important for our considerations here, what both rationalism and conventionalism or subjectivism share is the removal of normative knowledge from doubt and thus from rational scrutiny and learning. Rationalism renders the normative rationally self-certifying, and conventionalism and subjectivism exempt the normative from rational appraisal. This concordance reflects the influence of a third, equally venerable and pervasive principle: (3) the reasoning process itself must guarantee transcendence. There must be no possibility of the process failing when it is used correctly. Reason, that is, must defeat Cartesian or strong skepticism: doubt must be impossible. For that to happen reason must be constituted in a tool that satisfies three requirements: (R1) Normativity: It must be normative, i.e., capable of distinguishing legitimate understanding from misconceptions (truth from falsehood, right from wrong); (R2) Imperfection-free: It must not introduce imperfections into its operation; and (R3) Transparency: That conditions R1 and R2 are met must itself be beyond doubt, i.e., must be infallibly recognizable.19 Within this framework, logic looks like the ideal expression of reason and, with formal maximum expected utility (MEU) for pragmatic decisions, like reason’s only expression. So formalism seems to be the metarational ideal account, indeed essentially the only account, of rationality. These conclusions seem to follow because of the prima facie clear and intuitive way logic-based and MEU formalisms meet the three requirements for an adequate theory of reason. Complementing this formalist framework, each formalist position is backed by a corresponding metaphilosophy. Central requirements for logical

The Problems Generalized, Diagnosed, and Reframed 

63

empiricism are: priority, the claim that metaphilosophical and philosophical doctrines belong to a First Philosophy that is logically and epistemologically prior to, and normative for, science (and hence all knowledge); formality, the requirement that all the major philosophical concepts and doctrines are ultimately reducible to specifications of logical structure; foundationalism, the claim that there is an indubitable basis from which to construct knowledge (namely, observation and formal reason); and antiskeptical certainty, which is now seen to follow. Empiricists and rationalists fully share this metaphilosophy. Under this framework the ideal of reason is the state reached on (ideal) completion of the transcendence process, a condition of having a complete set of guaranteed truths. (This includes truth-tracking inference, since every valid argument is equivalent to a logical tautology.) Achieving the ideal of knowledge is also having a complete set of guaranteed truths. Thus the two ideals are equivalent. (This is no surprise because both are derived from logic; it is simply a sign that the issue of normative knowledge is removed from rational scrutiny under these conditions.) This formalist account of reason does provide a program for pursuing the ideal of reason. In fact, formalist assumptions are so deeply built in to traditional analytical philosophy that their acceptance too often goes unnoticed.20 However, given the multiple difficulties of the formalist approach noted above, another, richer program for specifying rationality is needed, one that will subsume formal reasoning as a small, but important, component.21 And there is room to move outside the confines of the formalist options. For instance, between the rationalist and conventionalist/subjectivist poles lies a range of constructivist positions where reason plays its role in the reconstruction of normative ideals, as well as in direct problem solving and decision making, as part of the learning process. Developing this possibility places normative knowledge on a par with other knowledge. Although doing so rejects the transcendentalism of the dominant Western tradition, this option is not incoherent, and it is supported by everything we know about the historical development of normative systems and about the general ignorance and limitations of humans, and by an evolutionary perspective on the origins and fallibility of our intelligence. From this perspective, and following Karl Popper, the demand for certainty inherent in traditional skepticism must be rejected as inaccessible to finite, fallible creatures like ourselves, and with that are rejected content foundationalism as well as requirements R2 and R3 on a system of reason (above). (Eternal transcendental truths are, one is tempted to think, the last gasp of a prescientific world.) This fundamental reorientation rejects the dominant Western ideal of reason as obtaining complete, certain knowledge.

64 

Chapter 3

Popper provides an argued case for bold conjectures subjected to rigorous testing as superior scientific method and correlatively adopts maximal informativeness as an alternative epistemic value to guaranteed truth, in turn opening the way to accepting multiple salient epistemic values.22 Popper abandons the requirements of epistemological foundations and certainty, as he must do given his critique of observation and his highrisk falsification method, substituting for them the requirement that epistemology explain how progress in understanding is possible (trivial for empiricism, crucial but substantive for fallibilism). This in turn prompts a double shift away from guaranteed truth as the direct, substantive ideal of science. The first shift is to a Kantian reconception of the ideal, where it is understood as inaccessible in itself but holds open the permanent possibility of improvement; it becomes a normative goal toward which progress can be partially and fallibly made and to which progress adds value. The second shift is, correlatively, to a multivalued conception of epistemic progress and, with values often in mutual conflict, measured in terms of partial improvements in many surrogate epistemic values. All this provides a rich framework that permits and rationalizes context-dependent trade-off among values in local methodologies, enabling the surrogate structure to partially and fallibly evolve over time as method itself develops. Moreover, to combat the misleading attractiveness of the formalist tradition, it is essential to keep in mind that the negation of each of the metaphilosophical doctrines of that tradition flows naturally from the alternative finitist, fallibilist perspective, where all claims are treated as limited and fallible and so potentially improvable through learning. Such positions permit a naturalism that is antitranscendental and prodevelopmental in its fallibilist learning approach to knowledge, both normative and substantive, while retaining a strong sense of normativity, i.e., without degenerating into conventionalism. This critique and its accompanying naturalist alternative do not require any objection to the centrality of reason to intelligence or the conception of reason as focusing on transcendence of imperfections, the first two core principles of the Western tradition. It accepts these principles but gives them a naturalist reading. Similarly, the principle that transcendence is effected through contact with an external reality is accepted, but now, that reality is the natural cosmos for acquisition of both normative and substantive knowledge and there is no empiricist requirement of infallible access. It is also important to be clear that formal procedures are not rejected out of hand; there are many situations in which it is eminently reasonable to use formal procedures and even many cases in which a refusal to use an

The Problems Generalized, Diagnosed, and Reframed 

65

available formalism would be irrational. The point is that rationality is neither exhaustively nor fundamentally captured by formal procedures. We are required now to take a new look at the nature of reason, the ideal of reason, and the program for reason. So far we have garnered several fundamental keys to constructing a more expansive conception of reason. First, there are the negative outcomes of rejecting • priority, formality, and foundationalism as metaphilosophical principles, • strong skepticism and the requirements of perfection and transparency for constitutive tools of reason, thus • the Western ideal of reason as complete guaranteed truth, and • formalism as an adequate account of the nature of reason and knowledge. Second, there are the positive outcomes of accepting • the Western conception of reason as central to intelligence, with the goal of transcending imperfections, • that rationality evinces a genuinely normative capacity, with the normative a fundamental feature of the natural world, • a Kantian account of ideals, with surrogates for those ideals that are finitely and fallibly but feasibly pursuable, • the interactive development of norms, methods, and empirical knowledge, • the fundamentalness of constructive development where reason plays its role in the reconstruction of everything, including substantive ideals and its own ideal, and • the fundamentalness of nonformal judgment and its skilled improvement as underpinning these constructive processes. 3.3  Naturalism Frames a Conception of Reason Suitable to the Pursuit of Science, Ethics, and Life in General The framework for freeing up a richer conception of reason proceeds from a shift to a naturalism of finite, fallible agents.23 Humans are part of the great evolutionary tapestry of life, differing in their capacities from other organisms by degree, not metaphysical kind, and often inferior in various ways (to eagles in sight, and so on). The improvement that reason provides is to be construed as a (partial) transcendence of imperfections within natural life, not a transcendence of natural life. Most of the radical implications of naturalism for developing an account of rationality derive from two encompassing features of organisms described

66 

Chapter 3

by science: fallibility and finitude. Section 3.2 has already noted the radical implications of thoroughgoing fallibilism, in particular, the explicit rejection of priority and foundationalism, the skeptical assumption, the requirements of perfection and transparency, and the formalist ideal of reason. Dually and constructively, fallibilism directly motivates (a) the adoption of a Kantian account of ideals with surrogates—but all fallibly learnable through the interactive production and improvement of norms, methods, and empirical knowledge—and (b) the fundamentalness of rational constructive development as in fact the learning process through which we pursue, and improve, our ideals and surrogates . Before examining the equally radical implications of finitude, it is useful to briefly review the general nature of a thoroughgoing naturalism. Broadly, the term naturalism as we use it refers to the position that the whole of life forms a natural (but not necessarily simple) unity, i.e., a single natural order or natural system, however complexly its parts are internally interrelated.24 An important immediate consequence is that only divisions supported by empirical evidence are to be introduced. Divisions imposed by human fiat are rejected; for instance, a priori metaphysical distinctions between mind and nature, linguistic and nonlinguistic intelligence, normativity and descriptiveness, and reason and cause are rejected. Naturalism is committed to treating humans as a natural species and hence to understanding human capacities, including the capacities for reason, knowledge, and ethics, as natural capacities that have evolved biologically from more basic capacities.25 This commitment imposes obvious constraints on the acceptable construction of theory. In particular, although philosophical theories are more general than scientific theories, the construction of a philosophical theory is considered part of the general epistemic process of understanding our world. Thus philosophizing yields fallible theories that are on an epistemic par with scientific theories. Albeit more indirectly, they are justified by the same kinds of explanatory considerations, and, up to a reasonable critical reserve, they must cohere with science.26 Human beings, like all other creatures, are imperfect: ignorant, as a result of their finitude, and in error, as a result of both ignorance and fallibility. Moreover, these imperfections run deep. Humans are born ignorant, not only of what the world is like but of what their own capacities are, and ignorant of what knowing and being rational (and loving) are and of what their pursuit involves. Everything more sophisticated than elementary responses has to be learned.27 Human error involves a wide variety of biases, for example, prejudice, egocentrism, anthropocentrism, and anthropomorphism, and pervasive inaccuracies, reflected in the failure of actions to

The Problems Generalized, Diagnosed, and Reframed 

67

satisfy goals and in inadequate beliefs and norms that motivate and direct inappropriate actions. In this condition human developmental tasks are central and urgent, but constrained and complicated by the pervasiveness of human finitude and fallibility. To be relevant, normative theories have to take the imperfections and limitations of human beings as central to what they prescribe. This realistic qualification applies, in particular, to the theory of rationality. As we shall see, however, this qualification does not preclude creating and striving for ideals. Rather, it requires that the standards that guide such striving must be attainable by the beings we are (since ought implies can). At the same time, those standards must also allow for cognitive processes capable of being guided by ideals and formulated in terms of improvement in the attainment of those ideals. From this perspective reason pertains to the developmental process, broadly understood to include all constructive self-change throughout life: reason is the organizational capacity to efficaciously prosecute imperfection-reducing / improvement-enhancing development.28 Rationality is located in the character of the improvement process itself. The formalist account of rationality, in contrast, locates rationality in the character of the end product produced (certain truth). In our adaptable, but finite and fallible, condition, however, we have no specific pregiven ideals as goals, and we are not equipped to recognize transcendental conditions like certain truth, so such a conception of reason is inaccessible to us. Reason regulates how we go, including pro tem constructions of our anticipated ends. Moreover, a focus on end-product in our conception of reason prevents recognition of learning about, and developing the structures of, rationality itself as we go, but this self-improving learning is crucial to being rational (see chapter 4). It is the reflection in philosophy of resolving deep problems, discussed earlier. What it is to be rational, or ethical, are deep problems that have to be clarified and reunderstood as we proceed, and similarly for all the questions that make up philosophy. Like the baby learning to walk, as our journey proceeds successfully, our control over and understanding of our going also improves. Reason as logic purports to deliver variously guaranteed truths, and fallibilist naturalist accounts cannot because they focus instead on the (fallible) process. In fact, trying to characterize reason naturalistically in terms of its outcomes is an inherent mistake since ends can no more be foreseen than can theories, and it is self-defeating since, as Popper made clear, aiming at security (read end-product reliability) and at informativeness are in conflict, a security emphasis missing the deeper, riskier insights while too strong a focus on risk-taking neglects assessment of what we can reasonably

68 

Chapter 3

rely on as a basis for further exploration. Practically, confining method to learning only from failures or to learning only from successes are equally too slow and too risky for learning; finite, fallible organisms learn best from mixtures of failure and success. Ultimately, its practical success is the only way we can naturalistically (but fallibly) judge the adequacy of either a proposed specific solution for a problem or a proposed rational procedure for solving a problem. Rational practices should normally confer improved success; if they do not do so, and given that errors in theory and data are reasonably shown to be absent, we have reason to doubt the rationality of the practices. Improvement in pursuit of other goals is not just the goal of reason; it also is the lever for improving reason itself. But how is the success of a rational practice recognized? Like other features of developmental processes, success cannot be fully characterized in advance, only in retrospect. At any time what counts as local success is simply satisfying the values delivered by the operative norms. The architect of a bank branch will look for design proposals that increase the value to the bank, whether in terms of short-term profitability or occupational health and safety, or longer-term client reach, market share and visibility, scale of client enterprise, and so on. But goals might differ over the course of the problem solving, perhaps changing the values deemed relevant (e.g., as public services or “green” principles are incorporated into the bank design), or shifting their priorities (e.g., greenness becomes a more desirable goal). As long as such changes in goals and values occur as problem solving proceeds and they increase the ranges of possibilities explored, evaluated, and accumulated as successively rich guides to next steps, they form one of the chief indicators that a powerful problem-solving process is ongoing.29 Since such changes often cannot be anticipated and might only emerge during the course of the problem-solving process, global or overall success in increasing the satisfaction of values cannot be specified in advance. Naturalism is adopted because it is fundamental to our most successful way of understanding our world so far—science. The history of science since the Renaissance reveals that progress has been made by naturalizing what was previously taken to be nonnatural. From the rejection of the heavens as a nonnatural sphere of perfection, which accompanied the rise of modern mechanics, to the discovery that the heart is a blood pump rather than a nonnatural center of feeling, which accompanied the rise of modern medicine, scientific progress has occurred as prevailing anthropomorphic and anthropocentric projections of nonnatural agencies into the world have been displaced by understandings of ourselves and our world as part of a single natural order. A stubborn holdout is the insistence that

The Problems Generalized, Diagnosed, and Reframed 

69

naturalized normativity is not possible. The theory of rationality developed here displays the same naturalizing spirit that has brought success to science, yet it claims to exhibit a genuine normativity. That conception of normativity must, of course, represent normative claims as being as fallible and learnable as the empirical claims of science. But instead of being the unthinkable repudiation of the obvious transcendence of norms that traditional philosophy would claim, it is simply part of the ongoing naturalistic revision of our self-understanding as we achieve a better grasp of the nature of the world and of ourselves as beings in that world. Hilary Putnam’s attempt to argue that reason cannot be naturalized provides useful support for this approach.30 Putnam proposes a set of four conditions or constraints on any acceptable account of reason and supposes that together they rule out any naturalized rationality. But any naturalist norm will satisfy these constraints, so we are free to adopt them, and not only for reason but also for ethics. The conditions are these: (C1) Reason is irreducibly normative; it is neither eliminable nor reducible to science. (C2) Reason is an autonomous capacity; it cannot be defined in terms of other normative capacities. (C3) Reason is to be conceptualized as a regulative ideal among other ideals such as truth, an ideal that we are permanently exploring but that we never fully attain; the exploration of reason is open-ended. (C4) At any given time, however, the specification of reason is context bound, the relevant contexts being those of our cultures, traditions, practices, and procedures. However, a naturalist account of any norm (a fortiori rationality) will take the form of a theory of the nature of the norm and of the distinctive normative role involved, a theory that will be fallible and revisable, the one that best captures our current, fallibly learned knowledge about that normative dimension and its pursuit. But it then directly follows that such a theory of any norm will satisfy C1, C2, and C4. And a Kantian account of the theorized norm, already adopted here, satisfies C3.31 What emerges from this discussion of naturalism and Putnam’s arguments are the following lessons for developing a naturalistic account of rationality: • It is necessary to develop a general naturalistic account of normativity, and of ideals in particular, together with a theory of the particular nature of reason, including the performance standards against which the rationality of natural agents is to be judged. This should be a fallibilist account that treats all these concepts as theoretical constructs in a fallible, learnable theory of our normative capacities.

70 

Chapter 3

• To complete the satisfaction of C1 through C4, reason should be modeled, not in terms of specifying the character of the products of rational process, but as an open-ended organizational capacity focused on the process of improvement, i.e., on supporting the pursuit of all ideals, including an ideal of reason. These principles are reinforced by the consequences of finitude. Formalist rationality theory stipulates that, to be rational, agents must formulate beliefs and values that are globally consistent and closed under deduction, take all factors into account when deciding, and so on. Not surprisingly, real people do not and, more to the point, in principle cannot check all their beliefs and values for mutual consistency, or hold them all simultaneously accessible for decision making, or determine all their implications for belief and action, or, typically, even perform all the desirable actions they can deduce from their beliefs and values. These “in principle” limits are remarkably strong; for example, any formal system that is undecidable cannot systematically review all inferences, and so forth, and this covers nearly all formal systems (a few of the very simplest excepted); and while we manage thousands of beliefs, it is not possible to check them for consistency, since even 40 or so would require more than the life of the universe to check for consistency using truth tables.32 Moreover, pragmatic versions of these constraints bite still more tightly; for example, whenever the anticipated costs of pursuing consistency and the like are higher than the expected benefits, it is irrational to pursue these formalist requirements, and such situations are found from facing charging lions to selling shares on the stock exchange, from resolving children’s squabbles to resolving adults’ marital conflicts. The general purport of these lessons is that, for any finite creature, satisfying formal requirements does not stand at the center of being rational; rather the judicious choice of smart heuristics does, where judiciousness is a matter of judgment informed by experience and acquired expertise, current evidence, and an appraisal of current circumstances, including idiosyncratic aspects. Think here again of the women’s decision making after genetic counseling (chapter 1). As Christopher Cherniak rightly explains, novel circumstances “are not peculiar to exotic scientific research; every child’s cognitive development involves a similar type of cutting edge of knowledge, where incomplete mastery of new domains ensures unpredictability. In these areas what counts as ‘reasonable care’ just cannot be well defined” (1986: 120). This common plight poses the core problem for intelligent procedure: how are we to solve problems in new situations or new domains where issues of

The Problems Generalized, Diagnosed, and Reframed 

71

what concepts, criteria, and procedures to use are themselves ill-defined and undetermined?33 Throughout this book we both inveigh against and utilize simplified models of human agents. Contrary to appearance, this is a coherent, if occasionally delicate, position to maintain. The kinds of simplification we inveigh against are those that obscure the foundations of things. In our case, these have to do with assuming that agents can achieve logical consistency, make all deductions from a body of premises, survey all premises to obtain those relevant to a decision, try out all relevant trial solutions to problems, and the like. All such aims are in principle unachievable by any finite agent, and their adoption serves only to obscure the actual basis on which finite human agents tackle problematic situations, including ethical problems.34 The kind of simplification we practice in this work is to leave aside all those complications that clutter, but do not fundamentally change, the model of human agency being used, and so could be “added back in” on occasions, and in degrees, required to explain agent behavior. These latter simplifications fall into two kinds. (1) Imperfections in the exercise of rationality: for example, biases and prejudices, illusions, delusions, and preclusions.35 It would be nicer to be without these, and some can be corrected through training to some degree, but most are empirically recalcitrant. But we can rationally discover these imperfections and learn how to correct for them, from eyeglasses to “data cleaning,” and in this very process exhibit the true power of our self-improving rational capacities. (2) Complexities of our inner emotional and cognitive development and self-regulation that mean that we often do not make fully conscious, or fully balanced, or fully pursued, or fully understood decisions.36 There will be plenty of situations where such factors play potentially important roles, for example, where doctor-patient relationships have both medical and developmental dimensions. Nonetheless, for the sake of simplicity we assume hereafter that none of the imperfections and complexities make an important difference to our general account and our basic understandings of examples. At this point, providing an adequate naturalist account of norms will likely be a pressing issue for some; however, its pursuit would also take us away from our focus on ethics, so we set it aside here. For interested readers the basic ideas have been laid out elsewhere.37 We close by emphasizing the process nature of reason—the rationality ideal is that of an ideal rational process, namely, to increase without limit the acquiring of relevant observational data, the canvassing of relevant reasons, the creative construction of relevant structures and processes, the subjecting of all to critical appraisal,

72 

Chapter 3

the improving of the skilled judgments involved in doing all this, and the improving of rational processes that doing these things affords.38 This process ideal of reason satisfies the core components of the Western tradition of reason: the centrality of reason to intelligence and the focus of reason on improving imperfections. However, unlike the degenerate formalist ideal, it rejects the Western ideal of reason, and the skeptical assumption. Instead, it is the right kind of rationality ideal for finite, imperfect creatures. This leads naturally to a rationality performance standard for finite agents: that rational agents try hard enough to carry out the reasoning tasks salient to the situation and feasible within their salient limitations in those circumstances. Here “trying hard enough” might mean context-dependently making all those inferences and/or consistency checks, using all those heuristics, and so on, that jointly contribute to increasing the feasibly achievable net benefit. This, for example, is a good overall characterization of the behavior of the women after receiving genetic counseling, and although we do not know enough about each woman to say how far she satisfied the required performance standard, it properly leaves open the assessment of inadequacies that might remain. 3.4  Conclusion When it comes to the Western metaphilosophical tradition concerning rationality and normative knowledge, we have (section 3.2.2), first, the negative outcomes of rejecting • priority, formality, and foundationalism as metaphilosophical principles, • strong skepticism and the requirements of perfection and transparency for constitutive tools of reason, thus • the Western ideal of reason as complete guaranteed truth, and • formalism (logicism) as an adequate account of the nature of reason and knowledge; and, second, the positive outcomes of accepting • the Western conception of reason as central to intelligence, with the goal of transcending imperfections, • that rationality evinces a genuinely normative capacity as a fundamental feature of the natural world, • a Kantian account of ideals, with surrogates for those ideals that are finitely and fallibly pursued, • the interactive development of norms, methods, and empirical knowledge,

The Problems Generalized, Diagnosed, and Reframed 

73

• the fundamentalness of constructive development where reason plays its role in the reconstruction of everything, including substantive ideals and its own ideal, and • the fundamentalness of nonformal judgment and its skilled improvement as underpinning these constructive processes. What emerges from section 3.3 are the following lessons for developing a naturalistic account of rationality: • The need to develop a general naturalistic account of normativity, and of ideals in particular, together with a theory of the particular nature of reason. This should be a fallibilist account, developed in interaction with other knowledge, formal, empirical, and normative, not held aloof from it, that treats all these concepts as theoretical constructs in a fallible, learnable theory of our normative capacities. • To complete the satisfaction of Putnam’s conditions C1 through C4, reason should be modeled not in terms of specifying the character of the products of rational process, but as an open-ended organizational capacity focused on the process of improvement, i.e., on supporting the pursuit of all ideals, including an ideal of reason. The normative performance requirements imposed on agents must, to satisfy ought-implies-can, be feasibly achievable by those agents. Although normative ideals such as the rationality ideal cannot be attained, the performance standards for those ideals must be attainable. The logic-based rationality conditions (i) fail to capture the essence of being rational, and so are inappropriate as an ideal of rationality, and (ii) fail to provide achievable conditions, and so are inappropriate for a performance standard. • The primary challenge for intelligent procedure is how to solve deep problems where the nature of the problem and what concepts, criteria, and procedures to use to solve it are ill defined and in need of resolution. Formalism is inherently incapable of responding to this challenge, yet deep problems and their practically successful resolutions occur whenever any creature masters surviving and thriving in a new domain. • The capacities needed to act rationally are substantially nonformal and constructive (a) because they involve not only choosing among rational strategies but choosing among context-shaping options, including the shaping of problems and their outcomes; and (b) more generally, because our norms, including those of rationality itself, and our specific strategies all have to be learned as an essential part of the process of improving ourselves and our circumstances.

74 

Chapter 3

The logic-based rationality standard locates the rationality of agents in capacities such as deductive closure and effective determination of unlimited consistency. Consistency, deductive closure, and the like are properties that a completed account of knowledge is expected to have. The antifinitude assumptions of logic-based rationality assimilate the criteria for ideal knowledge and the criteria for ideal rationality, and that collapse goes unnoticed, thus unargued, because both are assumed to be end products. In its complete deference to outcomes that are necessarily correct, hence rational, logic-based rationality ignores the key structural features of the rational processes of finite agents listed in the lessons above. Whereas requiring consistency, for instance, is legitimate on all occasions (e.g., even when facing a lion it is never good to conclude that one should run both east and west) and there will be contexts where its pursuit is the appropriate primary activity (e.g., in mathematical proof construction), the unconditional pursuit of consistency is not rational for finite agents, so it cannot be conceptually fundamental to explaining rational processes. Yet process, not product, is central to the rationality of finite agents. Rational process regards the skilled making of judgments about matters such as choice of context and heuristics as fundamental to the rationality of finite agents. Because it has no way to account for the rationality of judgment, the logic-based performance standard entirely suppresses judgment. In doing so logic-based rationality theory hides the more expansive conception of reason required to understand the rationality of finite agents. The posited unlimited capacities of logic-based rationality do not elide just some fundamental attributes of the improvement process—for example, context-dependency and heuristics—but the entire process, leaving only a presumed end product. An important distinction also is lost—the distinction between a process that enables us to (partially) transcend, i.e., reduce, our imperfections, and the ideals toward which improvement moves us closer, for example, knowledge and truth. With this preparatory understanding of the need for a more adequate, more realistic account of rationality, we now can turn to the task of developing a naturalistic account of reason for finite, fallible agents.

4  The Problems Resolved in a New Account of Reason

C

The discussions that culminate in chapter 3 are not purely critical; they also provide creative insights into the kind of account of rationality that is needed to rationalize the actual capacities for learning and problem solving that humans, with all their finitude and imperfections, in fact have. In this chapter we construct a broad, general account of rational processes; one that responds to the challenges and insights raised so far and that suffices for our purposes here, which concern ethics. Developing the relationship of the central core of this rationality with risk taking, reasoning in its various modes (e.g., counterfactual, probabilistic, modal, and causal), and the like, on the one hand, and, on the other, responding to the manifold imperfections in reasoning that cognitive science has in fact shown, are each large and proper undertakings, we acknowledge; but they are for another day. 4.1  Skilled Deliberative Judgment The foundational capacity underwriting our rationality is the exercise of skilled deliberative judgment.1 Every step in problem solving requires judgments. To illustrate this claim and motivate its adoption, consider the following essential steps in problem solving and the kinds of judgments they involve, in each case illustrated in the women seeking genetic counseling (chapter 1). (1) Establish the relevant normative and factual aspects of problems. The women refused the implied normative/descriptive division offered by the specialists and chose to include a much wider range of norms concerned with identity, family, and beyond. (2) Determine how these aspects are properly understood and assessed in tandem. For instance, the women needed to determine what was fair for the child given the child’s potential health burdens and how they might cope as parents. (3) Resolve conflicts among norms or values through constructive compromise. The women needed to understand how securing family support

T

76 

Chapter 4

might conflict with risking and raising a seriously affected child, and what compromises between them could be accepted. (4) Identify options, their accessibility conditions, and their potential contribution to solutions. In the case of the women, for instance, there were various levels of support (or resistance) to consider from partner and other children, siblings, and their families, social and for-profit services, and what costs (monetary and not) each might involve. (5) Identify relevant methods for pursuing options. The women’s methods could range from determining what value has highest priority (e.g., with respect to meeting a child’s physical needs), compromising among several values (e.g., among personal time caring, pursuit of income, and family indebtedness), and using MEU (e.g., when considering use of various purchased services). (6) Identify where problem, solution, and method require alteration to improve access to increasing value of solutions. The women’s shift from providing their personal values to an MEU, to instead exploring postnatal scenarios of caring for themselves and their immediate and extended families represents a coordinated transformation of problem formulation, methods, data required, operative constraints, and solution formulation that provided access to richer, more insightful and more resilient solutions. And this short list of kinds of judgment barely scratches the surface of the judgments tacitly involved in our approach to even everyday problem solving; for example, nothing has been said about judgments of reliability (for instance, by the women of themselves post birth). None of these judgments, or any other judgments, can be replaced by formal methods. Judgments are fundamental to processes of deliberation. Skilled judgment, which has to be learned and is fallible in its exercise, is valuable and ubiquitous. Although no judgment is guaranteed to be correct, it is far better, for example, to have a certified, experienced airline pilot assess the weather conditions for flying rather than a passenger. Fortunately, judgment can be improved systematically through learning and training. The method adopted here is that of making and improving skilled practical judgments, which Harold Brown uses in his examination of rationality (1988). It is this learnable improvement that makes skilled judgment an appropriate basic tool for rationality because our deepest rationality lies in it. And because, thanks to higher-order learning, improving skilled judgment is itself a skilled process open to learned improvement, rationality itself can be improved over time—we see such improvement in the increasingly powerful scientific method, in the spread of ethical resolution processes in institutions, and so on. Accounting auditors, who are faced with making professional judgments about the treatment of tax, investment, risk, and a multitude of

The Problems Resolved in a New Account of Reason 

77

other commercial affairs, many with clear ethical implications, have found deliberative judgment to be central to their skills, and there are accounting institutes devoted to training to improve skilled judgment.2 Indeed, one’s current skilled judgment rarely suffices for rationality, except momentarily, because current knowledge, acquired through use of those skills, is always feeding back to improve the skills themselves and to expand the situations in which skills can be applied. In the case of sensory perception, for example, the investigation of perception has both revealed the huge array of biases, distortions, illusions, and the like to which our perceptual processes are subject and also, in intimate interrelation with that, produced important improvements in, and corrections to, its functioning and a vast instrumented expansion of its capacities (microscopes—optical and electron; telescopes—radio, optical, X-ray; etc.). It is in this respect a useful model for the whole development of our understanding of rationality. Thus, it is rational to improve substantive judgments through rationality, and at the same time to improve the skilled procedural judgments that constitute the exercise of rationality. While everyone can, and should, improve their skills, finitude and diversity of life situations will mean that inequities of skill and experience have to be recognized as an important and ubiquitous feature of life. Rational decision making is then often improved by having a variety of other experienced persons (both lay and expert) play appropriate modulating roles in making decisions, from ethics committees to governments.3 This introduces in a natural way a necessary social component to being rational. It becomes obvious in the case of science, with its elaborated social processes of disciplines, research teams, and peer review, and is nowadays becoming equally obvious in ethics, with its elaborated processes in corporate and medical ethics, legal proceedings, governmental commissions, and so on. For the traditional approach of logic-based rationality theory, the situation is reversed: reason eliminates judgment in order to secure certainty of truth. This suppresses both its individual and social roles. Of course, none of the procedural issues just outlined go away; they are simply suppressed by assumption of agent infallible perception, unlimited logical powers, and so on. These assumptions unfortunately deflect from acceptance of two fundamental agent capacities—agent finitude and fallibility—which so limit accomplishment. Finitude and fallibility also limit the ability that rescues us from those limitations, namely the capacity for higher-order learning while engaged in problem solving. Humans have nonetheless a distinctive history of successful problem solving, and this despite the problem solving processes resisting attempts at reduction to rules.4 What counts as a good

78 

Chapter 4

next jazz improvisation/ surgical procedure/ chess move/ research design/ medical care decision, for example, depends on what the previous improvisations/ patient history/ chess moves/ researches/ medical responses were and cannot be generated by any manageably usable rule set.5,6 This situation comports best with skilled judgment processes, that is, with deliberation, as the expression of the basic character of rationality. At this point an elephant in the room must be identified and stared down. The foregoing discussion develops a case that humans do in fact rely on skilled judgment throughout their lives as the basis for their rational capacities, but it does not tell us how that reliance could be justified epistemically. If the response is that they do so on the basis of skilled judgment, it looks circular, with appeal to arbitrary subjective conviction lurking in the background as its dismal fate. Yet before we despair, note that the situation is in fact the same for judgment about formal systems. Appeal to intuitive apprehension of a priori truths and the like is not only beyond naturalist reach, it is in the same difficulty as the appeal to judgment: there is nothing stronger to safeguard our judgments from error than more judgment. Thought otherwise should be arrested by the following. Incisive Descartes had no better answer than the notion of ideas that were so clear and distinct we just could not be mistaken about them, but categorizing which ideas are sufficiently clear and distinct requires a set of judgments. This is typical of the fate of all such proposals. Nor have we even always been right in our judgments, even in logic.7 In practice correctness is judged by weight of judgment, from expert to crowd. This allows for social and historical critical evaluation, slowly rooting out errors—in fact by the very same means to be elucidated below for judgments. The power of these means should not be underestimated; they have allowed us to discover and develop powerful rational tools, from logics and mathematics to learning theory, perception theory, all manner of informing instruments, from glasses to microscopes, lathes to lasers, and more, and to correct all manner of imperfections in these in bringing them to their present, fallible, but mature forms. In short, we do not have, and in practice have not needed, a more surely founded basis for rational agreement than what we have for skilled judgments. 4.2  Four Resources for Rationality In the exercise of reason, there are four principal resources or means for the improvement of skilled judgment: observation, the use of both formal and nonformal reasoning procedures, constrained but creative construction, and systematic critical appraisal. These four bundles of processes are

The Problems Resolved in a New Account of Reason 

79

utilized both by individuals and by communal groupings of various kinds: in science by individual scientists, laboratory teams, learned societies, collaborative and discussion groups, universities, etc., and more generally by the lay public at large to various degrees in various domains. And in ethics similarly by ethicists, formal institutional ethics committees, and less formal procedures and training committees and groups, universities, etc., and more generally by the lay public at large to various degrees in various domains. Each of these four judgment-improving resources (observation, etc.) generates a set of constraints on rational procedure, and these constraints mutually interact in complex ways. For instance, the requirement for systematic critical appraisal may lead to the introduction of new persons to an evaluation process (scientific or ethical), persons who synergistically also add to the variety of observations brought to bear on the situation. Together these constraints constitute both disabling constraints (ruling out some possible actions as not rational) and also enabling constraints (making some actions rational that were hitherto inaccessible or inappropriate), and together these two sets of constraints structure and guide rational processes. In these ways, each of these means of improving judgment will contribute to both the elaboration of reasons for, and critical challenges to, our judgments—leading to improved judgments, for example, about the formulation of a problem, the method to be used to tackle it, and the proposed trial solutions to be pursued. And this process will also play various roles in developing our rational skills further. That they are the only kinds of judgment-improving resources available is itself a fallible judgment subject to reexamination as our understanding of reason improves. To elucidate the operation of the four judgment-improving resources, we briefly explore here how these resources severally contribute to rational processes. 4.2.1  Observation Entering a room quickly, a person notices an unexpected movement and sees it as a black cat jumping off a chair, only to find upon closer inspection that it was a black scarf blowing off the chair in the draft from the opening door. The initial perceptual judgment was wrong, but notice these points: (1) without observation no useful specific judgments about the room could be made; (2) the perceptual judgment was not wholly wrong—there was a movement of something—and that judgment provided a useful starting point for further enquiry; (3) the error in the initial perceptual judgment can be corrected by further, relevant observation; (4) organized individual observation can be more effectively used to improve perceptual judgments about what is going on in the room than can disorganized or solely

80 

Chapter 4

individual observation. This latter is because organized observation can focus the search on: (a) acquiring missing relevant perceptual judgments (can the object be located again?), (b) rechecking previously acquired perceptual judgments (does it still look like a cat?), and (c) arranging for the most suitable perceptual judgment-forming conditions (obtain more light to see properly? have other observers look from different locations?). These points extend to all observations. In particular, they extend to ethically salient observation. As much as for judging their medical needs, medical practitioners learn, or should learn, to observe their patients in order to judge their ethically relevant needs. For instance, they observe to discover if a patient is in denial, or trying to avoid responsibility, or instead anxious to be told everything so as to feel more in control, or concerned about managing a difficult family situation. An important part of becoming an ethically experienced practitioner is developing such observational skills so as to treat each patient during the course of consultation with care for their overall well-being as they manage their illness in their particular situation. Especially given the sometimes deep interrelations between a patient’s clinically relevant condition and prognosis and their condition as it pertains to overall well-being, ethically relevant skilled observation is not only inherently appropriate but often an important component of best clinical practice. In short, observation processes can be organized to improve the accuracy of perceptual judgment, and that provides one means of transcending subjective points of view. Further, socially organized observation can be used to distinctively, and typically more powerfully, improve individual perceptual judgments because socially organized observation (1) can implement observation processes that are inherently unavailable to a single individual, for example, simultaneous observation from distinct locations, (2) can be more discriminating than individual perceptual processes, either by including more specifically discriminating individuals and/or by focusing more observational resources on a problem than any individual could provide, for example, a large search party, and (3) is more able to correct errors of perceptual judgment by using agreement among observers and rechecking as required. Moreover, socially organized training of individuals in observation can improve the power and accuracy of individual observation, as when an artist or a scientist is trained to observe the subtle relevancies that escape the untrained eye. The result is a virtuous cycle in which both individual perceptual judgments and social organization of judgment making are mutually improved, along with accumulating understanding of how to go about

The Problems Resolved in a New Account of Reason 

81

making still further improvements. All of these processes are intensely combined in the regulation of perceptual judgments in science. All of them are equally available to support ethical judgment. Clearly, for instance, socially well-organized observation can improve ethical decision making by illuminating more aspects of situations and correcting individual observer biases, and the methods for arranging this can themselves be learned and improved. Child care workers concerned with ethical choices about parenting learn both what to observe about families in order to assess their parenting difficulties and options and when to call in others (a mentor, or psychologist) to check or supplement their observing. One of the side benefits of a well-functioning institutional ethics committee process is that it can make such resources available in a socially organized way in medical care settings. The unique property of sensory observation is that it brings us into intransigent contact with the external world in a way that is sufficiently independent of current beliefs and practices to be epistemically useful. In particular, it is capable of robust agreement or conflict with current theoretical expectations, including those concerning its own explanation. And it is basic: absent some transcendental insight, physical interaction with our environment (plus the internal equivalent in kinesthetic and like signals) constitutes our only intransigent access to our world, so its epistemically useful functioning is a sine qua non for both science and ethics. (This much is correct in the traditional views: observation plays a special role in science and in the factual characterization of ethical situations.) The job of rational experimental science is to increase continually the variety of useful observational interactions (e.g., by adding ultrasound probing to visual looking), their reach across kinds of situations and states of nature (e.g., telescopes extend our seeing into distant space and past cosmic states), their sensitivity to informative signals (e.g., neural probes can now detect the electrical activity of individual neurons), the precision of their outcomes (e.g., compare micrometers with rulers in length measurements), the range of their detection (e.g., a microphone can detect sounds both too high and too low in frequency for humans to hear), and their functional reliability (e.g., a spectrometer is a more reliable detector of light frequency than is unaided human vision). The underlying generic form of this formula is one that will apply to improvements in all four judgmentimproving resource domains. In return, improved observation provides for improved scientific theories, including improved scientific methods. This is a core part of the upward spiral of mutually reinforcing scientific improvement.

82 

Chapter 4

That improvement also contributes to improved ethical informedness. It is easy to forget, for instance, that perception of distress and denial is an ability as much in need of improvement as any other observation of states, and that it is a complex procedural skill. And, as with observations in science, such improvements feed directly into improved ethical treatment and also feed back indirectly into improved observation via new standards of assessment and the addition of new aspects to be observed (e.g., of dietary regime or of family resources for dealing with conflict). The job of rational applied ethics is to increase continually the variety of these useful interactions (e.g., by adding conversational protocols to passive observation); their reach across kinds of situations and states of nature (e.g., adding observation in familial settings to that used in clinical settings); their sensitivity to informative signals (e.g., the verbal cues that a patient is distressed or feels excluded from participation); the precision of their outcomes (e.g., the many subtle ways in which behavior can indicate pain in dementia patients who can no longer communicate verbally); the range of their detection (e.g., game playing of various kinds can provide clues to disturbed states in young children); and their functional reliability (e.g., agreement between a medical practitioner and a social worker may be a more reliable detector of patient depression or anxiety than either’s unaided observation). Perception brings two additional advantages with it: (1) Perceptual interaction takes place all our waking time no matter what we are doing (and some important processes continue during sleep). While this has obvious survival advantage, an epistemic advantage is that we may use perception in a significantly undirected, curiosity-satisfying mode, just to “find out what is there.” Whatever its ultimate rewards, curiosity is a core motivation driving our exercise of rationality in both science and ethics, for example, in exploring new data domains and creating new instruments of observation and in critically assessing trial outcomes from various angles. (2) Perception also has the epistemic advantage of generating conceptual content. Our creative intellects are dominated by the use of metaphor, and the primary source of metaphors is perceptual experience of interaction.8 But while individual perception is fundamental, it is but one of many physical interactions that can be utilized to construct information about the state and behavior of some system. Instruments of scientific observation and conversational protocols are other examples. All these interaction processes share the same basic features. (a) They are limited to a particular interaction modality, such as sound waves, and each has a characteristic frequency range within which it can register changes and beyond which it cannot (e.g., microphones extend human hearing, short interviews of

The Problems Resolved in a New Account of Reason 

83

children elicit higher quality of information concerning childhood trauma). (b) Being dynamical physical processes, they yield outputs to which multiple interactions, both salient and extraneous signals, distorting processes, insensitivities, malfunctions, and so on may contribute. (c) They transform their basic physical output into terms relevant for the larger investigatory process within which they operate. Natural perception involves a brain process akin to theory construction; in many ways the outcome perception is a plausible construction from our sensory stimulation, couched in concepts suitable for subsequent (inter)action (recall the earlier black cat/scarf example). Scientific and ethical observations are similarly constructed in terms of the expected theoretical conception of the situation in which the interaction takes place, whether what’s observed is a high-energy particle collision or a patient’s emotional response. (d) The observation performed through any of these processes (scientific or ethical) is subject to improvement by one or more of the following methods: (i) adding prostheses that correct or extend perception (e.g., lens to correct astigmatism, interview protocol to collect systematic data, etc.), (ii) specifying transformations of the raw data to represent particular features (e.g., angular momentum, depression) and to extract the relevant information from noise and bias (e.g., various statistical correlation analyses), and (iii) specifying the conditions for reliable observation (e.g., for humans, good light and no distractions, sympathetic and nondictatorial audience; for instruments, temperature range, incident vibrational frequencies and amplitudes). In consequence, no observation process delivers simple, direct truths about the external world. To be epistemically useful, what observation yields must be subject to manifold critical assessments (judgments) concerning all of the above aspects. In other words, the data have to be interpreted.9 Each of the complex theoretical and practical factors making up the observing interaction offers a corresponding locus of potential investigation, for instance, in responding to an observation that is judged anomalous (i.e., that conflicts with other observations or with a well-supported expectation). Conversely, though, each of these same factors is supported by whichever successful theories and practices represent the current scientific understanding of it (whether physical or psychological); and they in turn have multiple further connections. For example, for scientific visual observation the further connections extend from practical lens design through neuropsychology to theoretical statistics, and for ethically salient observation they extend through psychology to sociology and through interview protocol design to theoretical statistics. All of this together determines the potential range of epistemic ramifications to supposing the anomaly to

84 

Chapter 4

originate with the observation and their plausibility as the origin of an error; that is, it provides an array of plausible explanations of an anomalous observation. That richly structured, multidimensional array is the basis of an interpretational latitude to observation for that situation; it remains undecided, with varying degrees of plausibility, whether observations are accurate or in various kinds of error, and there is the latitude to appeal to any one of these options if they would best fit with the results of other critical examinations. That set of open possibilities is set inside another, broader set concerned with wider potential inadequacies. For instance, even if all its component processes operate correctly in themselves, observation may be so constrained by practices and expectations derived from beyond the observing context that it is incapable of uncovering the salient information—for example, because it is not undertaken in the relevant conditions to constitute a crucial test, or its output is incompletely processed, or it does not interact with the relevant features. Notoriously, for example, Galileo struggled to convince learned as well as lay people of his time that they could not observe absolute rest and motion, but only motions of one object relative to another. All the while he understood that unless the relativity of observed motion was accepted, simple observation would be taken to falsify his hypothesis of the motion of the earth around the sun instead of the reverse. When genetic specialists confine their observations of their patients to their family history of medical disorders and the results of any relevant laboratory tests, supposing this to suffice for understanding their rational decisions (see chapter 1), they have so blinkered their observations as to render them inadequate to their patients’ needs and, perhaps unwittingly, also made it impossible for either themselves or their patients to understand any alternative for their counseling roles or the use of broader factors in their decision making. The resolution of these latitudes in any given case is guided by method, theory, and practical skill, and by features of the specific context, all of which are themselves corrected and confirmed by refined observation. Here finitude requires that investigations be discontinued at the point where it is reasonably judged that the most salient issues have been sufficiently addressed for the moment.10 It is the business of our other rational tools for improving our judgments—reasoning, creative construction and critical appraisal—to try, along with the constraints issuing from observation, to have science and ethical practice progress while keeping our interpretational options within manageable proportions. So, while epistemically workable, the relation between the claims of science or practical ethics and

The Problems Resolved in a New Account of Reason 

85

observation is complex and nowhere epistemically certain in the manner empiricism hoped. Settled observations are better understood as the outcomes of a complex whole-of-science or whole-of-ethical-evaluation process, even while observation is also the basic informational resource for those processes. In this double manner we improve our own rational means of knowing the world, and, most important, we simultaneously improve our resources for making further improvements. We need to think of both science and ethics in these respects as dynamic systems, transforming their observing skills and capacities, and their use, as they develop. Judgment is required because of the absence of rules that are sufficient to decide such matters as how to develop an instrument and test its range, how to use it in the current circumstances, and how to interpret its output, not to mention broader matters such as whether to carry out an experiment or ethical intervention, when the experiment/intervention ends, and what response should be made to the outcome. Judicious judgment is required because all these aspects require coordinated critical assessment, in the face of finitude constraints. And expertise is required because only those expertly trained in the relevant disciplines and practices can make judgments judiciously on most of the issues.11 Like any other complex tool, observation can lead us astray as well as toward improved knowledge, ethics, and rationality. The roots of error lie within the design of each instrument as well as within our own limitations and imperfections. But, as our historical experience shows, this unavoidable risk does not prevent the development and use of improved tools and ought not to be allowed to suppress continued improvement. For when all of the specialized human and material resources are appropriately assembled (a matter of further judicious judgment), we generate the grand virtuous spirals of science and ethics, each ascending toward its ideal, improvements in each of observation, method, theory, and practice leading to improvements in the others, each step along this spiral constituted by a set of judgments, while each consequent improvement is used to improve our capacity for good judgment. This process itself will conform to reason if at each step the judgments that tentatively close discussion are reversible and all of the elements are in principle open to criticism and testing against alternative interpretations and observations, and if in each case these resources for reassessment have been applied to the maximum extent judged reasonable in the circumstances. In particular, and to repeat, the rational tasks of each of experimental science and practical ethics are to increase continually the variety of these useful observing interactions, their reach across kinds of situations

86 

Chapter 4

and states of nature, their sensitivity to informative signals, the precision of their outcomes, the range of their detection, and their functional reliability; and, to the maximum extent feasible, to bring the resulting observational resources to bear on the improvement of methods, theories, and practices. 4.2.2  Reasoning Methods Ignaz Semmelweis, a nineteenth-century physician, was concerned, both clinically and morally, about the high rates of death from puerperal (or childbed) fever in maternity wards, and particularly concerned that the newest, most medically advanced ward had a higher death rate than did older wards. He surmised that contamination with “cadaveric matter” may cause childbed fever and reasoned that if this was the primary cause then there should be a significant difference between women served by doctors with cadaver-contaminated hands compared with the patients of doctors who had sufficiently cleansed their hands. An experiment to test this prediction resulted in decreased deaths among patients in the hand-cleansing arm. But Semmelweis’s methods were either ignored by, or attracted general opposition from, the tradition-bound and/or self-serving doctors of his day. Semmelweis died at the age of 47, shortly after being committed to an insane asylum.12 Despite this unfortunate outcome, Semmelweis’s reasoning from his causal hypothesis was effective and essentially correct: if A causes B, then if A is present, B will occur (contamination arm), and if A is the major or only cause of B, then the absence of A will result in a marked decrease in B (hygiene arm). Mill codified simple causal reasoning of this kind in what are known as Mill’s methods; Semmelweis used what Mill called the “Method of Agreement and Difference.”13 The use of (generalized) Mill’s methods lies behind much experimental design. This is an example of how methods of reasoning provide constraints on the set of judgments we can accept, both epistemic and, in consequence, moral—constraints that are just as powerful as those provided by the judgments that arise from observation. In the case of Semmelweis, his hypothesis was eventually borne out, and improved hygiene led to a large decrease in hospital maternity deaths. The basket of formal reasoning tools of this sort includes at least formal logic, a codified set of rules for arguing validly, that is, in a way guaranteed to preserve truth; rules of causal reasoning (Mill); statistical analysis and inference; and risk-utility satisficing or maximization.14 Each of these forms of reasoning adds valuable tools for improving judgment. To illustrate and appreciate this last remark, consider deductive logic. Its major advantages are as follows. (1) Localization: each deduction is decomposable into small,

The Problems Resolved in a New Account of Reason 

87

simple individual steps that can be independently checked. (2) Partitionability: local reasoning about the valid deductive consequences of a small collection of propositions can be studied without reference to any wider set of beliefs and hypotheses to which they may be associated.15 (3) Simplification: logic greatly simplifies information by restricting it to 1/0, yes/no, true/false judgments; all the graded, partly overlapping, multidimensional, complex distinctions the world may demand (e.g., that between oblate and spherical, culpability and responsibility) are replaced by 1/0 artificial divisions offering only approximations to real relationships. Simple judgments of this kind are also more easily checked for formal correctness. (4) Context independence: valid reasoning is a context-independent, simplifying procedure and is useful for communication across distant contexts where shared substantive assumptions cannot be made. For all these practical reasons logic is an attractive reasoning tool for strongly finite agents like us.16 In science and ethics logic can be used to test individual theories and models for self-consistency and to test pairs of theories/models for mutual consistency; it can be used to derive predictions/obligations, falsifications, and explanations; hence it can also be used to reveal assumptions, by showing what extra premises would minimally be needed to obtain a valid derivation. These aspects form a fundamental methodological skeleton for science and ethics. All these are examples of how the use of formal reasoning increases the variety, reach, sensitivity, precision, and reliability of the arguments we can make in science and ethics, thus contributing to the development of both scientific and ethical theory/modeling and method. In these many ways, then, logical formal reasoning extends and refines our capacity for good scientific and ethical judgment. To this extent the traditional account of the role of logic in reason is correct (even if the rationale is very different from the traditional one). But, in keeping with section 3.2.1, and particularly when applied to persons, with their concerns for historical and functional holistic integrity, it is equally obvious that it has limited applicability and needs to be enriched by the other forms of formal reasoning, as well as underlain by good nonformal judgment. All formal reasoning methods also provide a means of transcending a local point of view, since everyone can use them and their conclusions are not user dependent. By widening and making systematic and public the demand for agreement about the consequences of methods, hypotheses, and practices, the use of formal reasoning again provides an important avenue for improving judgment in science and ethics. And, like observation, reasoning can be socially organized so as to transcend individual judgment and improve method, in both science and ethics, through (a)

88 

Chapter 4

implementing reasoning processes that are unavailable to most individuals, for example, complex computer programs, (b) being more discriminating than individual reasoning processes, for example, through employing many expertises, and (c) increasing creative deployment of argument and (d) critical scrutiny of claims and arguments through social communication and requiring suitable invariance of reasoning judgments across reasoners. All of these processes are intensely combined in the regulation of reasoning judgments in science, and can be just as intensely combined in ethical problem solving. As with observation, there is an ongoing reciprocal interaction between individuals and socially organized groups, in reasoning and improving reasoning, that can be powerfully mutually beneficial. Although when wrongly organized it can equally be damaging, requiring wise judgment to avoid, when productively organized the result is a virtuous cycle in which individual reasoning judgments and social organization of reasoning are mutually improved. But despite its value, the contribution made by formal reasoning to the improvement of judgment should not be overestimated. It is, for example, inherently incapable of supporting reasoning from evidence or from principles beyond simple deduction (section 3.2.1) and simply passes uncertainty from premises to conclusions. Reason by itself, like perception by itself, markedly underdetermines rational belief. And further limitations could be added, for example, the noncreative nature of formal inference impoverishes formalist cognitive theory, since the capacity to construct new concepts, principles, methods, and practices is central to human intelligence. And there is no natural place in logic for heuristics, for methods that relinquish logical guarantees of reaching their goal in favor of being simpler and faster and still offering better-than-chance odds of being roughly correct (such as deciding on a tree to escape a charging lion); yet for finite, fallible agents these methods are central to rational behavior. Another limitation is that, ironically, the creation of logic itself is clearly a matter of intelligent judgment, not formal rule following, for all nondecidable fragments. All this suggests that logic should be treated as a fallible theory, in principle like any other—a view supported by the deep relations between logic, algebra, and projective geometry that result in empirical dynamical theories exhibiting an underlying logical structure. This includes the nonclassical quantum logic, for which not all classical tautologies are quantum tautologies, suggesting that classical logic might actually be subtly false and at least that it has embedded empirical content. Logic might then be considered our most general theory of possibility.17 A rational resolution of these issues is to be pursued by the very process that we find in the extant

The Problems Resolved in a New Account of Reason 

89

literature: open critical debate enhanced by further creative construction of all the relevant formalisms and the continued observational testing of the scientific theories in question (such as the practical use of fuzzy logic on the Tokyo subway and the imminent rise of quantum computation). Time has been spent here on logic because it, more than other reasoning tools, needs situating in the larger space of naturalistic tools for improvement of judgment. After that examination, it might seem that the traditional alternative formalism for rational action, decision-theoretic utility maximization (or satisficing, for finite fallible agents18), is in a very different position. Utility maximization has no unbreakable rules (except that the maximization requirement be met); it is open to as many distinct values as there are recognized dimensions to utility, nor need those values be mutually coherent or immutable; and priorities among utilities can be highly context-dependent. Finally, and more radically, it allows the possibility that organized social practices may be emergent from multiagent interactions, including the systematic deliberative and behavioral practices of science, ethics, and law, to the extent that they exist. For these reasons, utility maximization offers a much more inclusive framework than logic for characterizing rational processes across many more kinds of situations in ways that provide satisfying explanations of capacities and behavior. For instance, the scientist can be modeled as a risky investor in scientific credit, yielding a model of risky belief commitment that replaces inductive logic in a way that gives explicit recognition to the diversity of inductive situations, indeed that applies in both normal and revolutionary settings. In general, this framework makes the multitude of diverse scientific practices—both those that mutually conflict and those that are simply different—explicable as efficient ways of spreading risky cognitive investment across institutionalized disciplines. It illuminates the conflicting nature of multiple epistemic goals, and context-dependent trade-offs among them, by modeling them as utilities while providing for the explicit recognition of the context-dependent priorities assigned to them and their capacity to develop over time as sciences mature. It even allows epistemic rules to emerge from these institutionalized activities much as price does from market trading, thereby providing a new model for understanding their origin and historical character. For all these reasons the decision-theoretic framework makes deep sense of institutional structure as an intrinsic part of the cognitive capacity and coherence of science and makes the design of that institutional structure itself a rational choice, learnably improvable. A decision-theoretic framework, in contrast to a logic-based analytic one, encompasses such considerations naturally and thereby radically expands

90 

Chapter 4

and enriches our conception of reason and its place in an understanding of intelligent action.19 But here too there are inherent limitations: maximization imposes huge demands on information and control of processes, so wherever these resources are scarce—for example, wherever there is deep uncertainty—it has no or little guidance to offer. For instance, it cannot handle situations where development of, and/or choice among, concepts, utilities, methods (e.g., investment rules) and decision contexts (e.g., games) play a central role in effective problem solving, because these actions either change the categorial structure of the decision problems or introduce radical uncertainty (or both).20 Neither does maximization help to understand the nature and limits of practical forms of reasoning such as metaphor/analogy, casuistry, and narrative analysis. Finally, its very broadness reflects a lack of guiding normative structure: utilities are given no constraints on their structure; maximizations (and thus economic efficiencies) are as multiple as specifications of system, time frame, and utility structure. All this argues in favor of regarding MEU models as relevant only after a problem has been otherwise properly structured, where they serve in the dependent role of providing a criterion for ensuring (sufficient) efficiency (that is, waste avoidance) when executing the problem solution. MEU modeling is ultimately a formal device, and its shortcomings largely derive from that. Thus, however valuable some of its insights, utility maximization modeling is best treated as but one formal tool among many that contribute to rational improvement processes. In sum, for both science and ethics a second crucial judgment-improving resource lies in our methods of reasoning: in both our formal methods—logic and mathematics, decision theory, and economics—and in our informal methods, including causal reasoning, reasoning by metaphor/ analogy, casuistry, and narrative analysis. These reasoning methods severally and jointly assist us to increase the variety, reach, sensitivity, precision, range, and reliability of the deliberations we are capable of. Among other uses, we can use deliberations to improve our testing and evaluation of our methods, theories, and practices—including, in return, our theories and practices of sound argument. 4.2.3  Creative Construction We have already seen an example of creative construction—the construction of reasoning formalisms like logic and mathematics—which at once makes clear the reality and importance of this category. Without the creative construction of reasoning formalisms, as the preceding section reveals,

The Problems Resolved in a New Account of Reason 

91

there would be no public frameworks for rational procedures, nothing to learn from concerning their limits or from which to attempt new creative expansions. The point generalizes. Consider the practical example of developing paintbrushes with synthetic hairs.21 Despite detailed copying of horsehair bend and the like, initial attempts at synthetic brushes all failed, the brushes delivering paint unevenly (lumpily) and in a rush. Sharper scrutiny eventually made explicit a hidden, unquestioned assumption about what horsehair features were relevant: a paintbrush is essentially a spreader, assembling a vertical layer of paint in front of it that it spreads over the surface as it pushes it along, essentially in the way soil or sand can be spread by dragging or pushing an upright shovel across a surface. But suppose this assumption was wrong, and instead a paintbrush is better conceived as a pump, with the hairs storing paint in the interstices between them by capillary action so that pressing down on the hairs as the brush moves pumps the stored paint out onto the surface. In that case it would be most important to reproduce the structure of the hair interstices and their spread under pressure. Success soon followed. This example well illustrates the centrality of creative construction—here a novel analogy—to problem solving. The analogical shift from spreading to pumping dramatically changes the fundamental suppositions about how the process proceeds and thereby redirects efforts to transfer the function to new materials, stimulating the development of new methods for interstice analysis, for hair design and construction, and for measurement of capillary actions, pumping pressures, etc. Analogical and, more widely, metaphorical thought is arguably the most fundamental mechanism of cognitive generalization, and we see how creative construction lies at its heart. It shows too that creative construction need not be explicit or primarily symbolically formulated (instead it might, as with pumping versus spreading, be grasped as practical motor behavior). The example also illustrates the other fundamental feature of cognitively directed creative construction: the importance of the contextual constraints guiding appropriate construction. In the synthetic paintbrush case the chief general constraints on a successful change of analogy were that it should explain how paint flow occurs and on that basis lead both to insight into why earlier attempts failed and to detailed improvements in brush performance, and do so while continuing to meet functional constraints: resulting brush not too large, too heavy, too specialized, too expensive, and so on. In science and ethics generally, creative construction is constrained by at least five broad classes of contextual judgments, namely, concerning

92 

Chapter 4

relevance/salience (for current problem, for “leading-edge” learning), cognitive resources (theory, methods, trial designs, data), epistemic capacity (performance requirements for a successful construction, specified as increases in epistemic values), ethical constraints (e.g., subject autonomy, public health, and safety), and practical constraints (financial, career, etc.). Sometimes access to superior solutions hinges on changing a constraint, for example, introducing futures trading to stock exchanges, which reduces producer risk while making markets more predictable for investors. In both science and ethics these constraints give cognitive form to creative processes; they channel it, and so shape it, and thereby help to make the creative effort more feasible. Constructions have to be creative if something genuinely new is to emerge, but if the overall process is not also sufficiently broadly guided, the costs of more open exploration can quickly outweigh any value to be gained. Constraints are also essential to our capacity to assess a construction: a particular creative construction may be an expression of an individual’s judgment, or that of a small group of individuals, but satisfying the constraints provides the means to test its quality in a manner that transcends the judgment of any individual or small group of individuals. Moreover, constructions are public entities and must ultimately be connected to public observations, arguments, and critical evaluations, bringing them under the other resources for rationality. In this stronger way, socially organized creative construction provides a way of transcending individual points of view while improving our knowledge and rational capacities.22 Problem solving within strong but not determining constraints has provided much or all of our most profound creativity, from symphonic composition and jazz innovation through architecture to fundamental science, and ethics also, for example, in crafting compromise and allocating kidneys for transplantation (see, respectively, chapters 6 and 9). Overall, practical ethics shows a healthy vigor in adapting to the many new situations our changing world presents, from the ethics of genetic alteration and counseling through ethics of design to ethics of corporate governance, and this despite having to function largely in the public societal realm with its diversities and crosscurrents. Creative construction, whether in science, ethics, or elsewhere, involves the creation of new concepts (e.g., transgender) and conceptual analyses (e.g., autonomy for the aged), new conceptualizations of problems (as in the case of the women who received genetic counseling), new analyses of assumptions (e.g., that autonomy is univocal), new theories or models (e.g., of doctor-patient relationships), new instruments and experimental procedures (e.g., novel interview instruments and protocols), new methods (e.g.,

The Problems Resolved in a New Account of Reason 

93

the equally partial but equal best analysis of ethical policy options and their implementation23), new institutional roles (e.g., new peer review and ethics committee processes), and so on. Without these innovations most of the power of our cognitive processes is lost; science and ethics would be crippled. In both domains, every sound experiment, every sound technological development, every sound piece of modeling, requires myriad intelligent creative solutions to problems posed by awkward constraints, incompatible performance requirements, the need to discover unknown details of processes or performance limits of instruments, checking for undiscovered errors, and so on. Without an ability to find creative ways through these problems, our scientific and our ethical advances would have long since slowed to a crawl. Instead, they are accelerating under a myriad of multiplying, creative synergies among all these features and more: institutional creativity—new laboratory designs and new institutional ethics roles, new journals, new collaborative research arrangements, etc., all supported by technological creativity—new kinds of devices, new procedures, new quality control methods (including ethics); and so on. Individual diversity in creative construction ensures a rich array of available constructions to consider, and that is essential to their productiveness of useful innovation, even while local sharing of expertises permits necessary cross-checking for errors and overlooked potential improvements. This applies to both science and ethics. In short, in both cases neither agreement nor difference is alone necessary or valuable; each is necessary and valuable in its proper role. A productive social organization has in fact a further subtle complexity: within their contextual constraints, creative constructions gain their distinctive normative powers from intricate, but systematic, patterns of difference and agreement across individual scientists or ethicists, for example, in ways that ensure that all the reasonably plausible alternatives for some problem solution are constructed and tested, along with selected “wild” constructions to test for present blind spots (as in the above example of early synthetic paintbrush hairs). In both cases constrained creative construction offers us new opportunities for organizing conceptualization, explanation, and investigation in ways that are sufficiently independent of current beliefs and practices as to provide deep or innovative learning. In the absence of transcendental insight, this has to be the way to open up our normative grasp, and to similarly transform our repertoire of physical interaction (including observation and communicative interaction) with our environment and each other and our methodological and explanatory resources. If these cannot expand synergistically with our expanding learning, then it is impossible to understand how finite, fallible

94 

Chapter 4

creatures beginning in ignorance can ever improve their learning either in science or in ethics.24 Pressing the diversity parallel between science and ethics raises the more general issue of the ultimate nature and source of normative rational, epistemic, and ethical diversities within and between human groupings. For nonnaturalist accounts of norms as a priori, disagreement can only be due to some kind of ignorance that nonetheless gives way to a priori certainty under some conditions. By contrast, once naturalists distinguish between what is appropriate to finite, fallible agents here now and those conditions that may be specified in unachievable ideal states (e.g., complete consistency), it remains everywhere open to learn about the most adequate normative concepts, principles, and methods in the process of empirically applying reason, science, and ethics. And the parallels ramify. Compromise among norms is unavoidable and ubiquitous for finite agents, hence it is so in both science and ethics. Recognition of diverse local cultures and skills is equally unavoidable and necessary for finite agents, for without them we would lose a special diversity, that of those local traditions that develop the highly sophisticated, specialized skills requiring years of devoted practice that are as crucial to quality science and ethics as they are to making music and musical instruments.25,26 Concomitantly, just as ethics and science share common general roles for diversity and unity within them, so too the diversity between them—for instance, of particular values (empirical accuracy for science, justice for ethics, etc.) and methods (empirical measurement of parameters for science, assessment of autonomy for ethics)—is compatible with their sharing common rational problem-solving processes, which is our contention in this book.27 It suffices for our purposes here that such creative construction processes exist, are improvable, and play an essential role in being rational. Creations are rationalized externally by satisfying their constraints deriving from the problem context and by their subsequent interrelations to the other judgment-improving resources for exercising rationality.28 To conclude with our version of the rational performance standard, it is required of rational science, and respectively of rational ethics, to increase continually, within feasible limits, the variety of normatively useful constructions (e.g., by adding statistical methods to logic, respectively adding institutional ethics committees to decision making), the reach of these constructions across kinds of situations and states of nature (e.g., statistical methods extending from macro to micro dynamics, respectively ethics committee responsibilities extending from family breakdown to corporate reporting and auditing), their sensitivity to informative signals (e.g., to single-neuron activity, respectively to

The Problems Resolved in a New Account of Reason 

95

discrimination of stress-resilient from non-resilient family dynamics), and so on. In return, improved construction provides for improved observation and improved methods. This is a core part of the mutually reinforcing upward spiral of normative and informative improvement. In contrast to formal reason’s narrowing of the scope of rationality to what is amenable to formal argument, the centrality of creative construction to rational process means that the scope of rationality is in principle widened to include all creative design insofar as it contributes to efficacious deliberation. It is not just scientific method, narrowly understood as logical inference that is a candidate for rationality, but the design of good scientific instruments, laboratory protocols, peer review processes, and so on. Similarly, rationality in ethics is not limited to the formal application of ethical rules, but includes the creation of ethics committees, clinical ethicist positions, bioethics courses and programs, as well as institutional design changes and legal responses and remedies, but also well-designed para-legal counseling and mediation, and so on. In short, the upshot of the application of reason in ethics is the design of increasingly sound deliberation, ultimately in support of human flourishing, just as the upshot of the application of reason in science is the design of increasingly sound deliberation, ultimately in support of the flourishing of human knowing. In both science and ethics we thereby achieve an enriched conception of rational intellect and a more satisfying grasp of our problem-solving endeavors. 4.2.4  Systematic Critical Assessment Discussion of each of the three preceding judgment-improving resources has highlighted the important role of critical assessment, in both science and ethics. For instance, observations cannot be taken for granted, given their limitations and error proneness, but must be supplemented, improved, rechecked for errors, and cross-checked for relevance and accuracy. This can be seen acutely in science in situations like that faced by Galileo where all the normal observations apparently supported the Aristotelian theory and it was what could not be perceived (absolute motion) that was the key insight. It also is acutely visible in ethics, in situations such as that of the women who received genetic counseling (chapter 1), where it was empirical information and value judgments that proved critical because the standard approach to moral decision making was repudiated. Critical assessment shades into improvement; for instance, organized observation, discussed earlier, forms an important part of the critical assessment process, for example, by acquiring missing relevant observations, rechecking previous observations, and arranging for the most suitable perceptual judgment-forming

96 

Chapter 4

conditions. And, as noted, logic, and methods of reasoning more generally, have an important role in critical assessment; for example, by being used to test for consistency, derive predictions, explanations, and falsifications, reveal assumptions as suppressed premises, and, when mathematically enriched, to construct theories of statistical inference for use in sampling methodologies, theories of error correction, methods for “cleaning” data, and so on. Critical assessment includes carrying out tests, proposing alternative explanations to challenge orthodoxies, checking performance and assumptions in wider ranges of conditions, improving the power of instruments and test methods, arguing the merits of alternative values and alternative assessments, and the like, and applies equally to both science and ethics (with some different orientations and emphases). Some of these activities can be performed by individuals, but, as we have seen, in most cases individual judgment is powerfully augmented by socially organized processes. The social dimension to critical assessment has a significance that goes beyond simply overcoming individual limitations and imperfections, for it has a role in problem solving that arises out of the conjunction of individual limitation and the power of diversity. Individuals cannot carry out all lines of inquiry (experimental, theoretical, or methodological) or acquire all expertises, pursue all frameworks and values, and so on; indeed, alone they often cannot carry out, acquire, or pursue even one. They must instead, perforce, carve out a sensible personal cognitive trajectory through this welter of alternatives, often with the only kind of coherent purpose available in this situation: to contribute to a superior combined effort.29 Moreover, there are many situations in which systematic observation and reasoning are not sufficient to reach agreement, because diverse individuals may decide to weigh the available evidence and/or relevant values differently, pursue different potential explanations for unresolved issues, accept different risk tolerances for closing off investigations, and so on.30 Irresolvable diversity of individual cognitive positions combined with individual fallibility makes critical evaluation necessary on two counts, equally important and interrelated. First, there is the necessity to check individual claims for error. Second, there is the necessity to integrate diverse individual claims so as to construct a coherent public scientific or ethical position. As already noted, this variety, arising precisely from the variety of roles finite individuals necessarily carve out, plays a crucial role in generating the range of alternative accounts (hypotheses, potential ethical resolutions), investigations (experiments and inquiries), methodological developments, and so on that constitutes the basis for a powerful learning process, scientific

The Problems Resolved in a New Account of Reason 

97

or ethical.31 At the individual level problem solvers demonstrate their rationality by arranging things so that, to the greatest extent possible, individuals follow their own commitments, within institutional constraints, while collectively they exhaust examination of the plausible resolution strategies. Collectively, the remarkable interdependency among individual values (epistemic or ethical) that such arrangements demand requires an institutional explanation focusing on the way in which problem solvers search for, and adapt to, the institutional roles they perceive themselves to play and the way in which these roles are designed in relation to the overall enterprise. Institutionally organized collective goals subsume individual goals but, rather than stifling variety, permit much greater variation among individual supporting roles. In this way science and ethics are inherently social enterprises characterized by fundamental, irresolvable diversity as well as by an extensible web of local and more global agreement.32 4.2.5  Synergy of the Four Judgment-Improving Resources The preceding discussions have anticipated a constructive dividend to the exercise of the four judgment-improving resources: their mutual improvement of each other. For instance, logic and reasoning generally will play an important role in the improvement of observation—for example, for science, by improving reasoning about the theory and practice of optics used in the design of microscopes and other visual aids; for ethics, in using anthropological observation methods to improve observation of relationships and interaction styles; and for both, in analyzing the data on individual perceptual performance and biases in laboratory and clinical settings. In the case of creative construction, tests for logical consistency will, for example, provide important information about, and usually form primary constraints on, new models and theories. And so on. These relationships are entirely mutual. For instance, creative construction contributes to improved logic through the construction of new formalisms, and the finding of counterinstances, often based on observation, can be used to criticize logical claims. Indeed such critical, creative appraisal of logical principles has stood at the heart of the historical process of improving formalizations of logic. It is actually in this way that the use of reason not only improves the content of science and ethics but also our capacity for reason itself. This is the essential bootstrapping synergy of all powerful learning noted above: the exercise of reason both improves judgment and the capacity to improve judgment. The idea that reason too should thus improve itself must sound strange if reason is thought to be absolute and known a priori, and may still sound strange if it is simply thought to be unitary. But once reason is accepted

98 

Chapter 4

as a fallible, learnable, internally complex organized procedural skill, with components that may each contribute to improving the others, the idea of self-improvement falls out naturally, and fits well with the social nature of rationality (as discussed later in this section). Of course, reason is also improvable through feedback from the attempt to apply it directly to the improvement of substantive judgment—for instance, as much has been learned about logic from the creation of theories of physics, computation, and expert knowledge as from mathematics, not to mention from the study of reasoning in psychology. The whole system of science and reason forms, at its best, a mutually improving synergistic process of great power. Exactly the same can be said for the relation of reason to ethics, and to social action generally. Indeed, it is the attempt to exercise reason in these areas, frustrated by restriction to logic, that spawned the development of casuistry, narrative, and analogical/metaphorical reasoning (e.g., in law) and other attempts to enrich our methods for improving judgment. To make explicit the underlying naturalism: this conception of an improving rational capacity brings theories and practice of rationality into the same learning dynamic as theories of science and of ethics, in general. The best evidence for deepening objective understanding is the emergence of a synergy of mutual improvement among all elements—skilled judgment and the four judgment-improving resources.33 Since marshaling the four judgment-improving resources for rationality requires expertise, and their effective marshaling is typically strongly context-dependent, there is no undifferentiated general individual ability to be rational, in either science or ethics. (There may be a general ability to learn to be rational.) And since there is wide individual variability in capacity, expertise, and role, plus variability in institutional design and organizational functioning, there is no undifferentiated general institutional ability to be rational either, in science or in ethics. Since expertise, both individual and institutionalized, is learned, neither is there a fixed rational capacity, given, say, at birth; rather, such capacities develop with the individual or institution they characterize. Moreover, the learning process is marked by finitude and fallibility. It is this interactive, open-ended, self-improving process that lies at the heart of being rational. And, as noted, it is an inherently socially structured process. Rationality requires the socially organized, systematic application of the four judgment-improving resources. Whence our actual capacities to pursue rational science and ethics depend upon judiciously exercising our capacity for apt institutional and cultural design, that is, for design that is cognitively effective and ethically sound. An isolated individual may be able to engage in

The Problems Resolved in a New Account of Reason 

99

some forms of utilizing each of the four judgment-improving resources for rationality, but such an individual suffers from genuine epistemic and ethical deficits arising from their finitude and fallibility. Both individual judgment and social organization are essential, and we do not reduce rationality to either alone. Succinctly: individual judgment without social organization is crippled, social organization without individual judgment is blind—but, as preceding discussions revealed, when acting together, mutually reinforcing individual and social capacities form the most powerful feasible frame for rational process. The social dimension is irresolvably complex for finite, fallible agents. As everywhere within culture, the relation of the individual to the group takes the form of a delicate shaped-and-shaping dynamic: in both science and ethics social process shapes individual development, but individual action also supports and transforms social process.34 Social goals and process norms subsume individual goals and norms, even while all social normative judgments are grounded in individual normative judgments. As well as whole-institution values and interests, all within-institution subgroups will have their own values and goals to some extent, based in different subinstitutional goals assigned them, competition for external resources and rewards, and so on. This shaped-and-shaping relationship includes many inherent tensions; for instance, the need for both diversity and expertise creates permanent tensions in those who seek to arrive at results in a rational manner. Diversity calls for many partial participants, most inevitably having less expertise, while expertise calls for fewer participants, with the benefit of long learning experiences; both are useful, but finitude insures they are mutually traded off.35 Resolving the tension is again a matter of judgment because, in advance of a known resolution of the problem at issue, any decision is fraught with risk and/or waste. Sensible choice processes will vary according to the specifics of the situation, and there will be no rules to follow. The discussions of skilled judgment and the four judgment-improving resources place us in a good position to offer a naturalistic conception of rationality for finite, fallible agents as an antidote to the formalist version. 4.3  Characterizing Rationality The essential project of reason in the Western tradition is transcendence of ignorance and error of all kinds, achieved through a process of reasoning about empirical experience that improves our state or condition with respect to our ideals and improves our ideals and their surrogates. Thus, for

100 

Chapter 4

Jean Piaget, reason just is the capacity to engage in the overall process of organized and organizing improvement, and that is also, broadly, how we understand it.36 This captures well a natural sense in which reason is central to intelligence. We begin with these questions: (Q1) what kind of account is to be given of a problem solution or decision being rational? (Q2) what is the source of content for an account of rationality? and (Q3) what should be the relationship of an account of rationality to an account of our empirical cognitive capacities? For the formalist the answers are immediate, respectively: rational solutions follow from logical inference from the initial premises (data, theories, etc.); logical inference is known by transcendental insight; rationality is the normative standard against which empirical knowledge is assessed. However, once we move to a finitist, fallibilist, naturalist approach, the answers are less obvious. (Q1) What kind of account of rationality? Remembering the conclusions reached in chapter 3, we see no alternative but to pursue a process account of rationality: what rationalizes an assertion is its undergoing a particular process and emerging from that process with appropriate credentials. Even formalism offers a quasi process: derivability from accepted knowledge. But an adequate response has to offer a process rich enough to encompass all of the diverse and complex aspects which we have seen are actually involved in problem solving and to bring to bear all feasibly available supporting resources. (Q2) What process with what resources? Following the preceding discussions, we characterize an initial account of rational process as one that provides for all feasible utilization of each of the four judgment-improving resources, along with the tools of compromise and reflective equilibrium formation (chapter 6), to improve judgment in problem solving and decision making. This core characterization suffices for our purposes here. But it is possible to briefly add a little generic elaboration of process. Following what was learned in chapter 3, the exercise of rationality takes this general form: (I) set up an initial formulation of the problem and of its anticipated solution(s) and the norms against which a satisfactory solution is to be judged; (II) judiciously bring to bear the four judgment-improving resources along with the methods of compromise and wide reflective equilibrium formation as appropriate; (III) formulate a proposed trial solution or partial solution and its attendant decisions; (IV) using (II), develop the proposal to the point where its adequacy can be tested against, for example, its ability, where apt, to generate empirically recognized improvement, its

The Problems Resolved in a New Account of Reason 

101

coherence with solutions/decisions made in other, related situations, and so on; (V) return to (II) with proposals that do not deliver sufficient value (perform sufficiently well), to be reworked against discovered sources of deficiency and productivity (failure and success) and applied to (I) and/or (III), and so on through the process until an outcome of sufficient value is found. Finally, add (VI): the quality of the decision-making setup has to be such that it activates the virtues (humility, cooperation, attentiveness, …) that together constitute what we might call the deep intelligence of the process, meaning that the design of the process exploits as fully as possible all of the psychological constructive resources that humans can bring to bear on problem solving and decision making. (Contrast here the initial and final conditions of the women who received genetic counseling.) Outcomes from a sufficiently rigorously applied process of this kind are considered rational, since there is no more that could feasibly have been done in the circumstances to improve the outcome. (Albeit the outcome and the process that produced it are fallible and open to improvement should that option arise.) To illustrate, consider brief sketches of how this process applies to science/ ethics/design, in that order. A problem solver (e.g., experimental researcher or hospital bioethicist or architectural designer), seeking to understand what is happening in some situation S, sets up an initial formulation of the problem and of anticipated solutions (what test best reveals S’s nature and how can it be realized? or what agreement best satisfies bioethical obligations in S and how are they to be implemented? or what design best satisfies client and public norms for S and how is it to be realized?). Then, judiciously bringing to bear the four judgment-improving resources plus the methods of compromise and wide reflective equilibrium and the deep virtues of investigation, our problem solver formulates a partial proposed solution—for instance, an experimental protocol or an ethically sensitive compromise or an architectural design theme—to investigate and, using these same resources, develops the proposal to the point where its adequacy can be tested against operative epistemic or ethical or client/public norms, including by diverse persons beyond the problem solver. If its performance is satisfactory (if it delivers sufficient value), carry it to completion as the solution. If not, reconsider the proposed problem and solution formulations, the method used and constraints assumed, and the proposed partial solution, using the same resources to identify sources of failure and success. Revisit these steps as often as is required to clearly identify sources of failure and success and thus obtain a sufficiently (epistemically or ethically or pragmatically) valuable outcome.

102 

Chapter 4

We devote the next chapter to reconsidering our earlier ethics cases to show the applicability of this approach. We note that, with (II) removed or held tacitly, this is essentially the widely accepted Dorst/Cross/Maher model of design problem solving process.37 However, the omission of (II) is crucial, as is the absence of higher-order learning in both accounts (see problem-solving model below). Nonetheless, the natural way this process fits science, ethics, and design supports the view that it is the universal core to all rational problem solving. We are not alone in thinking that problem solving might be a useful perspective from which to understand rational process in ethics. There is, for example, the insightful work of Caroline Whitbeck relating online engineering decision making to ethical decision making, to be examined in chapter 6. Kees Dorst and Lambèr Royakkers (2006) also suggest that problem solving in design theory could prove useful to understanding rational ethical decision making—rightly in our view, though they do not go beyond that suggestion to develop any substantive account of rational process the way we do in this book. (Q3) What should be the relationship of rational norms and processes to empirical cognitive capacities? Contrary to the formalist’s a priori dichotomy and asymmetry between norms and empirical fact, the normative naturalist holds that both the normative and the empirical accounts are fallible, learnable theories and as such have to be mutually adapted in a respectful two-way iterative adjustment. Adaptation of each theory is driven by success—the empirical theory is driven by success in understanding the actual problem-solving process humans use, both its strengths and its weaknesses, biases, and limitations; and the normative theory by success in guiding the investigation of problem solving, improving our investigatory capacities in the process (cf. Brown 1996). In consequence, the two theories are locked into an iterative developmental process in which improvement in one is used to improve the other. At any given time, our rationality norm—some version of the process I through VI—should reflect our current best cognitive conception of problem solving and consequent decision making, all inherent imperfections noted and compensated or quarantined. We can provide some promising support for this approach. Based on evidence from problem solving in scientific research and in design, and from the cognitive science of expertise, one of us (CH) has proposed a universal innovative problem-solving model that provides a common core process across all problem-solving domains. This process has the key property that it can solve (if anything can) even the initially ill-defined problems that necessarily lie at the basis of all problem-solving domains, from how to do

The Problems Resolved in a New Account of Reason 

103

valid research under fundamentally new conditions, how to design a new kind of object, and how to develop ethics for a new kind of domain, like genetic engineering, to how to develop a new formalism suitable for a class of problems, how to solve a crime, how to throw a good party, or how to become a saint. This core process turns out to be a particular form of the I through VI process described above, thus providing the cognitive counterpart to our rationalizing norm. Its main emendations are (a) to develop an account of changes in problem formulation, method, constraints, solution formulation and partial solution proposals as the main dimensions of change in problem solving and (b) to provide a parallel higher-order process of learning the general character of problem solving in that domain (by remembering, from specific problem solvings, which kinds of problem formulation, method, and so on fail and which succeed). Remembering patterns of success and patterns of failure is central to obtaining higherorder knowledge about problem solving in new problem domains and in turn explains our capacity to develop expertise, including for solving that domain’s initially ill-defined problems.38 We are in a position to adopt this rationalizing process in ethics, as both normative and reflecting our best empirical understanding of the rational process. 4.4  Conclusion The account of rationality developed here does not provide a finalized definition of rationality, much less any a priori or transcendent sense of normative status. Nonetheless, unlike other accounts that purport to do so, it respects each of the principles and lessons developed in chapter 3. It also unifies the complex cognitive dynamics of scientific and ethical development and provides a naturalist, yet noninstrumental, theory of rationality that is rich and powerful enough to illuminate development of rational capacity itself. It provides a proper basis for understanding the intratwined process of improving rational learning capacity, including self-improvement, while developing solutions to problems—a process central to solving problems. Consequently, its content, while not formally specified, is determined by this scope. Reflexively, this theory of rationality was itself constructed through the same rational process of applying the four judgment-improving resources using skilled judgment, thus it gains its support in the naturalist manner. In ensuing chapters we focus on its application to illuminate ethical decision making and development. On that score its theme of continuous improvement holds for ethics itself. The general conception of ethics presented inter alia in this chapter

104 

Chapter 4

is one of the design of flourishing living for all. That conception opens up the space of the applicability of ethics to encompass all dimensions of personal and public life, a conception that dovetails with correlative design processes everywhere, from urban architecture to political institutions. The scope for the improvement of ethics is nothing less than the scope for the improvement of design for flourishing across all these aspects.

5  The Process of Ethical Resolution: Using the Resources of Nonformal Reason

C

Science and ethics are seamlessly, reflexively theoretical and practical. In the preceding chapters we showed how restricting rationality to formal reason produces distorted, divided, and inadequate accounts of the theoretical and practical dimensions of science and ethics. By avoiding the dichotomies that formal reason creates, nonformal reason integrates theory and practice and thereby provides unified understandings that elucidate and enhance the rationality of science and ethics. In the rest of the book, we develop a broader, more constructive account of ethics that is parallel to the more faithful and realistic account of science that proceeds from nonformal reason. This chapter illustrates how the resources of nonformal reason are used, first by young patients seeking a resolution of a poignantly distressing problem and then as a way of responding to the deep moral challenges posed by several of the vignettes in chapter 2. 5.1  The Rationality and Morality of Dying Children The ethnographic research of Myra Bluebond-Langner (1978) on how children with leukemia discover, despite a conspiracy of adult silence, that they are dying and then deal with that information in a way that enables them and their parents to retain their dignity, identity, and humanity presents a surprising and illuminating example of the use of the resources of nonformal reason. A troubling moral problem for the parents of terminally ill children and for the health care professionals who care for them is whether to tell them they are dying. Often the children are not told. Why? One obvious answer is the fear that doing so will cause too much worry, anxiety, and distress, and parents want to protect their children as much as they still can. Another obvious answer is that the standard task-oriented rationales for telling the truth to terminally ill adults do not apply. Children do not have to give consent to treatment or to research protocols; they do

T

106 

Chapter 5

not have to put their financial affairs in order; and they do not aspire to repair relationships with alienated or estranged family members or friends. Given these differences, the potential benefits of telling might not outweigh the potential harms of telling. But a status-oriented moral reason might be offered in response to such a consequentialist analysis: children are persons, and simply by virtue of being persons, they have a right to know what is happening to them. Even if that claim is correct, the moral agonizing is moot for a practical reason: these children do not have to be told, because they know. That is what Bluebond-Langner discovered: “All of the leukemics that I studied knew their prognosis. All knew that they were dying before death was imminent” (Bluebond-Langner 1978: 165).1 Two mothers suspected as much, as an exchange between them reveals: Mrs. Smith:  It’s pretty hard to keep things from them, no matter how hard you try. Mrs. Andrews:  Have you told your son? Mrs. Smith:  He knows he has a blood disease, and, well, I don’t know. But they do pick up quick. (226)

In fact Bluebond-Langner found that children acquired a vast amount of information, not only about the hospital, the people who work in it, and how it is run, but also about their disease, its treatment, its progression, and its prognosis. And she found their acquisition of all that information “even more remarkable because they learned in a situation in which the parents and the staff unconsciously conspired to keep them in painless ignorance” (165). So how do these children “pick up quick”? They are eminently rational, but not in the manner formal reason prescribes or in a manner that formal reason can recognize or explain. Nonformal reason, however, provides an understanding of their remarkable accomplishments. 5.1.1  The Rational Discovery of the Diagnosis The children whom Bluebond-Langner studied were able to “pick up quick” not just because they are, like most children, curious but because they are, like most children, intelligent and rational. The rationality of these children resides not in what they learned but in how they learned, that is, in the processes they devised and used to discover their diagnoses and prognoses. The resources of nonformal reason figure prominently and essentially in those processes and in doing so illuminate their essential rationality. Observation  Much of what the children learned about their situations and their surroundings came from careful, intense observation. On their initial entrance to the oncology clinic to which they had been referred for

The Process of Ethical Resolution 

107

outpatient treatment following diagnosis, they became engrossed in watching children at play in the waiting room. They sat by themselves but had opportunities to observe and question those near them. During their initial clinic visits, Bluebond-Langner reports, “the children, like their parents, watched more than they talked” (176). In the hospital the children detected a hierarchy of authority among the physicians and were able to construct that hierarchy largely by observing the behavior of the attending physicians responsible for their care and the house staff (the residents and interns) (148–149). They noted other evidence as well. Resident doctors were differentiated because they were younger than the attending hematologists, who were in charge of the leukemic children, but older than the interns, and because, like the hematologists but unlike the interns, they generally wore lab coats over their street clothes. The information the children obtained from their observation not only allowed them to discern similarities and differences, it enabled them to draw helpful inferences. Once they recognized that the hematologists made the decisions about their care, for example, they could challenge a resident who came to perform a procedure by questioning whether the hematologist had authorized it, and after hearing their parents argue with a resident about a procedure, they could stall the procedure by making a fuss about it. The children’s acute observation operated more widely. By watching the behavior of residents, interns, and medical students in the evening, for instance, the children were able to identify, with remarkable accuracy, romances between staff members (151–152). And they inferred the likely functions of rooms in the hospital from observations of the equipment in the rooms (136). These children display an uncanny perspicacity. They know what to watch, how to sift from everything they watch the features of the world that are relevant to them, and, as we will see, how to critically assess their experiences and use them to refine and focus their subsequent observation, i.e., to improve their observing. From the information they obtain from their persistent, dedicated watching, they create classifications and posit generalizations, which they then use reasoning methods to test. Formal and Informal Reasoning Methods  The children employed familiar reasoning methods to understand their surroundings and what was happening and was likely to happen to them. To begin to comprehend who they were, they quickly differentiated themselves from children with other diseases: All the children could distinguish between leukemic and nonleukemic patients. “They [the leukemic patients] come to Monday clinic.” “We [leukemic patients]

108 

Chapter 5

all have the same blood disease. You know Greta, she comes [to the hospital] all the time. Well, she has a blood disease [sickle cell disease] too, but she goes to another clinic.” (155)

Reasoning methods enabled them to explore, develop, and test the significance of their observations. They used induction and analogies to draw inferences from their observations; for example, a boy’s loss of hair indicated that he was on the drug vincristine because other boys had lost their hair after being put on vincristine (103). Their reasoning often relied on assessments of other people’s behavior: “Alan … knew that things were not going well when, in spite of how good he felt, his mother would cry. ‘In my house, crying and bad things go together’” (180). Similarly, they made comparisons with other people’s past behavior to get a sense of the seriousness of their illness: The children put people’s behavior together with the tests and treatments they were receiving and concluded that they were really sick, “seriously ill,” “very, very sick.” “This is not like when I had my tonsils out”; or even “when I cut my head open.” “This” was somehow worse, as were they. (173)

The children appreciated that arguments by analogy have two sides: the relevant similarities between the items being compared that support the analogy and the relevant dissimilarities between the items being compared that challenge the analogy. When a death occurred or was imminent, the children were concerned with establishing the cause of the death. Having done that, they compared the deceased to themselves. Children who only recently had discovered their prognosis emphasized their dissimilarities to the deceased. But children who had known their prognosis for some time realized that their similarities to the deceased were more compelling (189). In short, they had learned how to improve their informal reasoning. At times the children used the deduction that typifies formal reason: Only children who are dying do not have to go to school anymore. All other children must go to school. I do not have to go to school anymore. Therefore, I am dying. (190)

Deduction, as we will see in the section on systematic critical assessment, is also the logical basis of the hypothesis testing the children did to evaluate the correctness of their inferences. Creative Construction  The children’s creative constructions often took typically childish forms. They would congregate in places where they

The Process of Ethical Resolution 

109

could talk privately, and they would devise strategies for eavesdropping on conversations. Older girls, for example, met in the women’s bathroom on the “adult” side of the clinic’s waiting room to talk about the disease, drugs, and what was happening to other children. When adults entered, they became silent and often dispersed. Aware of what was going on, boys gathered near the bathroom door (144). Because the children knew that information exchanged between adults was more valuable and more reliable than information adults conveyed to them, they surreptitiously sought that precious information. They would try to delay separations from their parents in the clinic’s waiting room so that they had more opportunities to overhear adult conversations; they would hide behind office doors; they would pretend to be asleep; they would turn down the volume on the TV; and they would ask people to be quiet when doctors assembled by the door of their room to talk to their parents. Sometimes they were more daring: one child planted a tape recorder in the doctors’ office, which served as a conference room. But one of Bluebond-Langner’s central findings—that the children proceeded through a series of stages of awareness in learning about their diagnosis and prognosis—shows a more sophisticated use of creative construction. She found that the children went through five stages in acquiring information: discovering first that “it” is a serious illness; then learning the names of drugs and side effects; grasping the purposes of treatments and procedures; coming to understand the disease as a series of relapses and remissions (minus death); and finally seeing the disease as a series of relapses and remissions (plus death) (166). At each stage the children had to formulate a new self-concept or definition of themselves: seriously ill; seriously ill and will get better; always ill and will get better; always ill and never will get better; and dying (terminally ill) (169). To move from one stage to the next, the children required both information and experience. When new information was forthcoming, their existing information had to be sufficient to enable them to recognize its relevance and incorporate it into a set of premises from which they could reason to a conclusion: Without the requisite information, the children could not integrate new information to come to a new conclusion. For example, if a child knew the names of all the drugs and their side effects, as well as the purposes of various treatments and procedures (stage 3), but did not know the disease was chronic (stage 4), the news of another’s death did not lead the child to conclude that the disease had a terminal prognosis. (168)

110 

Chapter 5

Experience both yielded new information and linked information: First, the children needed the disease experiences (e.g., nosebleeds, relapses, bone pain) to gather significant disease-related information. At any sign of illness, the children were taken to the clinic, where they could again meet their peers and discuss what was happening to them. Second, the disease experience enabled them to assimilate this information by relating what they saw and heard to their own experience. (168)

Bluebond-Langner found that children could remain at one stage for an extended time and that “age and intellectual ability were not related to the speed or completeness with which the children passed through the stages” (169). Consequently, some younger children with average intelligence knew more about their prognosis than some very intelligent, significantly older children who were still in their first remission and had made fewer clinic visits. The older children knew only that they had a serious illness. They needed more experience to construct the subsequent stages and realize that they were going to die (169). Creative construction is a third crucial component of this process because it gives the children the capacity to use their information and their experience to construct and evaluate new conceptions of themselves. They can do this because formal and informal reasoning methods enable them to abstract from the information they have acquired through observation, personal experience, and conversations. Then, systematic critical assessment allows them to test and revise the self-concepts they create until those self-concepts match the information they possess and explain the experiences they have had. Systematic Critical Assessment  The children conducted systematic critical assessment in two important, familiar ways: to assess the credibility of evidence and to test inferences drawn from that evidence. With respect to evidence, the children astutely realized that information is reliable only if it comes from trustworthy sources. So they tested people who served as their sources, including Bluebond-Langner: Benjamin asked everyone he saw that day what happened to Maria [who had just died]. Later, when I asked him why he asked everyone, he said, “The ones who tell me are my friends. I knew Maria died. I saw the cart come for her. They told everyone to go in their rooms. I wanted to see if you were really my friend.”2 (188, note 27)

Their critical scrutiny included, not surprisingly, the hospital staff: The staff’s refusal to answer questions directly, and parries such as, “Well, we’ll see” or “I don’t know,” eroded their credibility. “I asked three times when I could

The Process of Ethical Resolution 

111

go home and they said, ‘I don’t know.’ They saw the tests. They told me that’s why [to see when I could go] I had to have them.” If the staff did not know, who would? In the children’s eyes, the staff had no excuses for not knowing. (180–181)

A hierarchy of trustworthiness emerged for the children. They generally trusted other children because they expected to get straight answers from them; with respect to drugs and medical procedures, though, they excluded children without experience from their discussions, perhaps for reasons of efficiency but more likely because they deemed them unqualified. The only information they prized more highly than information from other children was information obtained by eavesdropping on adults because they knew that adults shared information they withheld from children. They considered information volunteered by adults the least reliable (180). The children tested the generalizations they inferred from their observations and conversations and the self-concepts they developed at each stage by employing a rudimentary form of scientific hypothesis testing. Bluebond-Langner describes a prevalent practice: The children tried out their newly acquired self-concept on everyone they saw. Children … would often exclaim, “I’m really pretty sick, you know,” and then offer as evidence the changes in people’s behavior towards them or the physical changes in themselves, especially the “wounds” suffered from procedures.… The children would begin by counting their needle pricks and pointing to them. (173)

At stage 1, Bluebond-Langner reports, children used this strategy to begin every encounter, and it continued to be used in the succeeding four stages, albeit less frequently. The repetition at stage 1 probably was important because the children were testing a very different—and undoubtedly frightening—conception of themselves. The outcome of the testing was that “the children found confirmed in people’s reaction to their pronouncement the fact that they differed from other children” (174). The same testing procedure was used to resolve other uncertainties. When a child died or was close to dying, for example, “the children attempted to establish the cause of death either by asking a question or stating a hypothesis and assessing the other party’s reaction” (189). In stage 3, after the first relapse made it clear that the disease was chronic, the children were, as Bluebond-Langner puts it, “left more and more to their own devices” (181). They had to obtain the information they wanted from other children, and to do so, “they formulated hypotheses about the relationship between various symptoms and the drugs, procedures, or treatments employed, and checked them out with peers” (181).

112 

Chapter 5

The testing procedure the children used is a version of orthodox scientific method: formulate a hypothesis, develop predictions from the hypothesis, and test the predictions to see whether they are confirmed or refuted. It is a drastically simplified and much less reliable version, but nevertheless an instance of a core strategy for generating knowledge in any natural realm. The children’s testing of self-concepts introduces another informal reasoning method: fit or coherence. The progressive growth of knowledge and the unfolding of the stages imposed constraints on how the children could respond to new information and experiences: Once children reached stage 4 … they had fitted all of their acquired information into a pattern of relapses and remissions, of never getting well. And just as all other events could be fitted into the scheme, so too could the death of another child. It could not be seen in isolation, but was part of the same chain of events that they were. They realized that they shared with peers not only the same experiences, but also the same prognosis. (184)

As they acquired more information, the children had to make judgments about how well that new information fit or cohered with their current overall understanding of their situation. When that new information was hard to reconcile with their prevailing understanding or exposed tensions or inconsistencies in that understanding, they had to make further judgments about how to create a new understanding that could accommodate it. The criterion of fit also operated more specifically. A child’s first relapse was profoundly distressing because the preceding months of remission had lured the family into thinking they could lead “normal” lives. With a relapse, the old symptoms were back, and the same medical procedures were being performed. Jarred from their comforting respite, the family had to drastically revise their feelings and their understanding of the situation. In particular, the child had to create a new self-concept that fit the new evidence. People’s reactions now were the same as they had been at the time of the initial diagnosis, except now they were crying more and telling less. In addition, the child was no longer getting the drugs for headache and bone pain that had been prescribed during the remission. Instead, the child was put back on the same drugs that had been administered after the diagnosis. To move to stage 3, the child had to create a self-concept that fit these discomfiting changes—and to give up the belief that the drugs had brought about a recovery. As the children used the resources of nonformal reason, their knowledge improved, and their capacity to use the resources to acquire knowledge improved. With respect to improving knowledge, for example, on their first

The Process of Ethical Resolution 

113

hospital visit the children immediately began to orient themselves as they stepped off the elevator, quickly making a crude distinction between “us” and “them.” “Them” were those in uniform, and “us” were those not in uniform (171). “Us” were the patients and families who took orders and spoke only when spoken to. “Them” came and went as they pleased and inflicted pain or delivered news that caused pain. Initially, the children were not aware of the criteria they used to make these distinctions. They recognized a multiplicity of differences, however, not just the obvious one between doctors and nurses, and they realized that the distinctions were not based solely on uniform. Over time they progressively refined the distinctions to develop an accurate understanding of the hospital staff and their functions. With respect to their methods, the children were able to critically assess their observational and reasoning skills, adapt them to new circumstances, and hone them as experience and necessity dictated. Bluebond-Langner reports that “as time went on the children became more astute, more capable of drawing conclusions from people’s behavior. They knew the indices for various types of behavior and what they meant” (173). The children appreciated how much they relied on observation and testing for information, and as they recognized changes in how people behaved toward them through the progression of their illness, they realized that their observational and testing strategies needed to be adapted and enhanced. In short, experience with using the resources of nonformal reason leads to their application to improving methods of extracting and applying information, making constructions and the like, which is how nonformal reason reflexively improves itself. This survey of how the children used the resources of nonformal reason highlights four important features of nonformal reason. First, these resources do not function singly and independently; rather, they operate jointly, reciprocally, and reflexively, and in ways that can improve their collective use. Second, the operation of nonformal reason is thoroughly grounded in experience. Watching others and reflecting on one’s own experiences are essential to acquiring and testing information and to the creative construction of self-concepts. Third, the operation of nonformal reason is robustly social. Talking, acting out together, and generally interacting and collaborating with others provide the children with crucial information, not just for making judgments about themselves and their peers in various stages of the disease process, but also for testing information and improving their methods of observation, creative construction, and systematic critical assessment. Fourth, the exercise of these skills is also self-improving.

114 

Chapter 5

The magnitude of what these children do should not be underestimated. They suddenly find themselves in strange and frightening new circumstances, indeed in circumstances so unprecedentedly different that they have to radically rethink and reorder their environment, their place in that environment, and, ultimately, themselves. They cannot rely on any of their old assumptions or rules, or on any of their old methods; everything has to be relearned. These children manage to do all that. How? Their challenge parallels that of scientists entering novel research environments—for example, quantum mechanical or ape language learning environments— where the basis of understanding is uncertain. In such new environments scientists cannot rely on any of their old assumptions or rules, or on any of their old methods, either; everything has to be relearned. Yet they likewise manage to do so. Again, how? Certainly not through just formal reasoning, which presents only paradoxes of circular assumptions and contradictions from which anything and everything can be deduced. Rather, they use judicious, iterative interaction among observation, creative construction, informal reasoning, and careful critical assessment. The same answer applies (albeit more modestly) to the children. 5.1.2  The Rational Morality of Dying Children Realizing that children with leukemia know they are dying, and understanding how they know they are dying, might seem to remove any residual moral responsibility to tell them they are dying. Yet it could be argued that the children should be told, both to confirm what they know and to give them an opportunity to talk about death. That position is theoretically attractive, but the lived morality of these dying children defies its abstract appeal. Bluebond-Langner observed the children’s “consistent impenetrable silence in the face of ultimate separation and loss” and sought an explanation of this perplexing phenomenon (198). The children are not silent because talking about death is taboo. Moreover, unlike adults who initially conceal a terminal illness but subsequently disclose that they are dying, the children’s silence is resolute. Why is secrecy so vital to them? It is vital because the practice of mutual pretense, which exists when each party knows the patient is dying but acts as if the patient were going to live,3 preserves the social roles and responsibilities that have acquired magnified importance for the children and their parents, enabling the children to retain their socially created identity as children and to continue to feel a sense of worth. A dying child fundamentally disrupts the social order, because the impending death of a child collapses the established roles of both children and parents (as well as medical professionals). Children are

The Process of Ethical Resolution 

115

supposed to prepare for, and be prepared for, the future, but a dying child has no future. Clearly the children understood their plight; when they knew they were going to die, they stopped talking about long-range plans and goals: They never again mentioned what they were going to be when they grew up, and became angry if anyone else did. A six-year-old boy was angered when his doctor tried to get him to submit to a procedure by explaining it and saying, “I thought you would understand. You once told me you wanted to be a doctor.” He screamed back, “I’m not going to be anything,” as he threw an empty syringe at her. She said, “OK, OK,” but a nurse present asked, “What are you going to be?” “A ghost,” he replied, and turned over. (194)

Yet practicing mutual pretense is difficult for all parties and sometimes requires quick thinking: When a seven-year-old boy’s Christmas presents arrived three weeks early and Santa came to visit him, suggesting that the boy would not live to Christmas, he turned to his choked-up family and the staff members around his bed, saying, “Santa has lots of children to see, he just came here first.” (204)

What makes the extraordinary vigilance and effort that mutual pretense demands of everybody worth it? In a society and a culture so thoroughly oriented to the future and so doggedly devoted to achievement, the prospect of death, in addition to the inherent fear and uncertainty it instills, erases a child’s identity and concomitant sense of worth. These children appreciate the hopes and expectations their parents harbor for them, and they know they will never fulfill those hopes and expectations. Adults go to work; children go to school. They no longer go to school or even do schoolwork. They will have no future achievements because they have no future. Childhood is about becoming, but there is nothing they will become. Those portentous losses give the children powerful motivation to hold on to as much as they can. Mutual pretense sustains their status as children so they still can be valued as children. Because the roles of children and parents are symbiotic, mutual pretense likewise sustains the status of their parents so they still can be valued as parents. The loss of authority and control over their children, not just in the hospital but in general, is threatening and debilitating to parents. They feel ignorant: “Many parents thought that they did not know very much, too little to answer their own questions, not to mention the children’s” (215). They feel overwhelmed and helpless: “The parents saw themselves as powerless before the disease, the doctors, the machines used for treatment,

116 

Chapter 5

and, most important, their children. They felt that they could not care for their children on a day-to-day basis, let alone in the future” (215). They lose some of the joys and pleasures of being parents. Mothers could not cuddle their children because being hugged or touched was painful, and most importantly, parents could not nurture, protect, and rear their children, the tasks that define parenthood. By preserving the established social framework, mutual pretense retains order in a world roiled and disordered by disease and thereby affirms the identity and worth of the children and their parents. As Bluebond-Langner emphasizes, “the social order is a moral order” (230). The children’s pretense is indeed deeply caring, loving, and altruistic. Bluebond-Langner comments, “Ironically, these children came to see their own task in life as supporting others” (232). An exchange with one child provides an unsurprisingly childish illustration: Myra:  Jeffrey, why do you always yell at your mother? Jeffrey:  Then she won’t miss me when I’m gone. (126)

The children’s gift to their parents was, however, also the greatest gift they could give to themselves. The children realized that sustaining the hopes of adults secured their continued presence, which meant that they would not be left alone (228–229). Mutual pretense was a way of satisfying the deep human need to belong: Everyone could retain their identity and membership in society, except those who did not practice mutual pretense; they were ostracized and abandoned. To put it another way, as long as individuals fulfill their obligations and responsibilities and do not violate the social order, they are granted continued membership in the society and all that comes with it—freedom from fear and abandonment. No one seemed to know this better than the terminally ill children; for they practiced mutual pretense unto death. (232)

The basic roles, structures, and institutions of society are designed to meet that fundamental need, but at the price of fulfilling the accompanying social responsibilities. In this regard, the loving, caring altruism of the dying children was supplemented and supported by an exquisite appreciation of their vulnerability. For greater than the fear of dying is the fear of dying alone, without comfort, without solace, and without love. The children did not want to be abandoned when they most needed to belong. Their arrival at a moral practice that is practically accessible in the circumstances yet meets the deepest needs of everyone involved is a model of constructive, creative ethical rationality at work; moreover, one that holds

The Process of Ethical Resolution 

117

its own among the adult achievements of civilized culture and society that support thriving personal lives. How, then, does the practice of mutual pretense emerge and, given the strenuous demands it imposes on all parties, persist? Conditions propitious to mutual pretense existed from the moment the children entered the hospital. The staff strove to withhold information from patients, to prevent leaks, and to cover up leaks and displayed an unwillingness to answer questions from the children by leaving the room, reprimanding the children, and ignoring their questions (200–201). Parents were reluctant to volunteer information, explained little, and displayed their unwillingness to communicate through lengthy absences from the room, brief interactions with their children, and avoidance of conversations about disease-related topics (201). The children observed this behavior and used informal reasoning techniques such as argument by analogy and induction to recognize the significance of the behavior, to construct patterns, and to infer the implications for their own behavior. Just as they would have countless times before, in their family setting and in the hospital, they recognized and interpreted the nonverbal signals now being sent to them—that adults were not willing to acknowledge their reality and were withdrawing from them—and they responded to those signals by fitting themselves into the changed family and hospital milieux. As Bluebond-Langner notes, “When they realized they were dying and that people were no more willing to talk about their condition than they had ever been, the children started to practice mutual pretense” (210). Although engaging in mutual pretense is a different kind of children’s pretendings, it was readily accessible to them because it was a consoling adaptation of their experiences with their parents and the hospital staff. If the process by which children arrive at pretense explains only its general character as an intelligent adaptation, the process of continuing pretense reveals more. Pretense is not a simple single action (deny I’m dying) that is repeated. Rather, pretense is a whole complex, contextual, and reciprocal pattern of relating that is needed to express denial and that must be continually assessed and reconstructed as circumstances change. How the children managed to maintain that pattern again displays their use of nonformal reason. The practice of mutual pretense involves following guidelines such as avoiding dangerous topics, keeping interactions brief, and ignoring emotional breakdowns, and the children devised tactics accordingly. Children who had never cried about painful procedures began to refuse the procedures by crying, in the way “normal” children do, instead of protesting, “Well, I’m dying anyway,” or “It won’t do any good” (204).

118 

Chapter 5

A boy and a staff member talked about past Christmas plans as if they were for an upcoming Christmas even though he was unlikely to celebrate it (202). As did their parents, children adopted a “distancing” stratagem to avoid or limit interactions with others. But because they were confined to their beds, the children could not avoid being in a room or make excuses to leave, so they had to refuse to talk, pretend to be asleep, lash out or cry out continually, or speak inaudibly or superficially (206–207). Whereas the children had formerly employed tactics for cultivating relationships, they now adopted tactics that would weaken and destroy relationships, to reduce the number of people with whom they had arduously to sustain the illusion that they were not dying. Preserving mutual pretense is, as Bluebond-Langner describes it, a “delicately balanced drama” (208). Preserving that precarious balance requires constant attentiveness and perceptiveness and the ability to make specific contextual judgments about competing demands. The children must be disciplined yet accommodating; committed yet flexible; strong yet sensitive; self-interested yet altruistic. Making good judgments involves the resources of nonformal reason: scrupulous observation of their surroundings and of others’ demeanor and reactions as well as their own demeanor and reactions; sound inferences about what has happened in the past and why it happened, what is likely to happen now, and what would be effective means to desired ends; creative responses to breakdowns and ways of preventing breakdowns; and unrelenting, comprehensive critical assessment of particular aspects of the practice (e.g., why did I slip up there, and how can I avoid doing that again?) and the practice as a whole (e.g., are people going along with me, or are they questioning and challenging me more?). The success of the children—that, as Bluebond-Langner reports, “they practiced mutual pretense unto death” (232)—testifies to their perspicacity. The children certainly were “willful” and “purposeful,” and their use of nonformal reason enabled them not only to act in the world created for them but to create their own world, the world that they and their parents needed. The justification of mutual pretense does not come from the application of a philosophical principle or theory. Mutual pretense is a complex, morally rich, socially embedded, life-affirming practice. No formal analysis could have deduced from unnuanced, blunt universal rules a position that subtly extracts moral value from ultimate loss through the subterfuge of a tacitly agreed lie. This is an unexpected, powerful—and wincingly beautiful—response that does moral justice to all parties. Only the whole intelligence, using all its informal resources, can arrive at so profound a

The Process of Ethical Resolution 

119

resolution. Nonformal reason is the capacity that makes such outcomes humanly possible. So, is practicing mutual pretense the right way to respond to the impending death of a child? That is a formal-reason question, particularly when asked without regard to the experiences of those involved in the practice or the diverse contexts within which the practice operates. Can the practice of mutual pretense be improved? Is there a better way to respond to the impending death of a child, perhaps a hybrid that preserves the identityforming, worth-creating roles of dying children and their parents yet allows the children to talk about dying? Those are the questions of nonformal reason, which is always critically attuned and searching for improvement. If there is a better way, it will not come from an abstract moral principle or moral theory. It will come from experience honed by careful observation and systematic critical assessment of and reasoning about mutual pretense, complemented with creative construction of novel alternatives. And ascertaining its superiority will be a matter of judgment. Above all, though, it will respect what the real rationality and morality of these children teach us.4 The dying children with whom Bluebond-Langner worked were richly rational and heroically moral. They were understandably eager and determined to find out what was wrong with them, and they used the resources of nonformal reason to acquire the knowledge they sought. They had an impressively mature and sophisticated appreciation of the impact of their illness on their parents, and they used the resources of nonformal reason to find a way of preserving the identity, dignity, and worth of their parents and themselves. They displayed seemingly preternatural moral acumen, sensitivity, and fortitude in maintaining, in the direst of personal circumstances, the silence of protection. They knew that the process of dying is for the living, and they knew how to guard the living.5 5.2  Using the Resources of Nonformal Reason in the Vignettes Chapter 1 presented women’s process of deciding whether to try to become pregnant after receiving genetic counseling as a paradigm of skilled deliberative judgment. In this section we draw upon selected vignettes from chapter 2 to illustrate how nonformal reason operates in various other processes of rational moral deliberation. 5.2.1  Ms. B The vignette about Ms. B (see vignette 2.1) was chosen to show how the meaning of crucial moral terms—in this case “euthanasia”—can be multiply

120 

Chapter 5

ambiguous, and that they cannot be intelligently applied until they are contextually disambiguated. The resources of formal reason are too meager for this task. How does nonformal reason manage it? Ms. B is afraid of dying by suffocation, so she has chosen terminal sedation as the means by which she will die. This choice seems clear to Ms. B and consistent with her Christian values until a family friend interprets the arrangement as euthanasia, creating confusion and dismay about whether her plan violates the religious and moral convictions of Ms. B and her family. The method of applied bioethics offers little help to Ms. B and her family, and little direction to Dr. M, because the concept of euthanasia is vague and indeterminate. Timothy Quill, Bernard Lo, and Dan Brock (1997) demonstrate that, in practice, various forms of ending life are hard to distinguish from one another, so determining which ones to categorize as instances of euthanasia is uncertain and controversial. In response to that difficulty, Quill, Lo, and Brock shift from a moral principle about euthanasia to a moral principle about respect for autonomy. For them, the core moral issue in endof-life decision making with competent patients is whether autonomy is respected, that is, whether the consent of the competent patient is voluntary and informed. But their alternative elides the real problem because Ms. B’s difficulty lies not in consenting—an expression of autonomy—but in the moral reflection and effective deliberation that give substance and meaning to personal authenticity, a fully realized sense of autonomy. To be sure, informed, voluntary consent has become the most important facet of autonomy.6 “‘Autonomy’ thus means freedom from external constraint and the presence of critical mental capacities such as understanding, intending, and voluntary decision-making capacity” (Beauchamp 2003: 24). But this is a narrow, thin construal of autonomy. Richer is Bruce Miller’s (1981) four senses of autonomy, discussed in chapter 2. Informed, voluntary consent is incorporated into two of the four senses—autonomy as free action and autonomy as effective deliberation. Requiring that a patient give permission for a proffered treatment in the absence of undue influence, manipulation, or coercion makes the patient’s decision voluntary; having information about a proffered treatment makes the patient’s decision intentional; and both together make the patient’s decision free. Awareness of the alternatives and the likely consequences of those alternatives allows a patient to evaluate options and come to a decision based on that evaluation, which is engaging in effective deliberation. Effective deliberation in this sense will be recognized as a weak form of the wider, more powerful deliberation process used by the women after receiving genetic counseling and fully supported by nonformal reason.

The Process of Ethical Resolution 

121

A crucial issue for autonomy construed as effective deliberation, however, is how effectiveness is determined. One way in which deliberation can fail to be effective is that the patient’s weighting of values can be nonrational, which, Miller explains, means “either that the weighting is inconsistent with other values that the patient holds or that there is good evidence that the patient will not persist in the weighting” (1981: 25). Although Ms. B’s problem is about values, it is not yet a problem of inconsistency between her values or of vacillation in adhering to her values. Until she was forced to, Ms. B probably never had envisaged that she would have to decide about the management of her own death. As well, Ms. B probably had never thought about euthanasia in more than vague, superficial terms, and when she did, she likely had never entertained the possibility that she would request euthanasia. When she had to decide how she would die, she assumed that not dying by suffocation is an intelligible, desirable, and morally permissible end and that terminal sedation is a morally innocuous means to that end. She believed that aggressive sedation would simply render her unconscious and thereby prevent her from experiencing the horrors of dying by suffocation; she did not appreciate that aggressive sedation also could depress her respiration and thus could cause her to die sooner than she would without it. Ms. B’s assumptions have been thoroughly roiled. Her goal of not dying by suffocation now is potentially constrained, if not precluded, by a norm that prohibits euthanasia. Because her situation has become complicated, confusing, and uncertain, she needs a thorough, comprehensive, critical examination of her plight. In that regard, Ms. B’s problem engages the two deeper, more complex senses of autonomy that Miller distinguishes: autonomy as authenticity and autonomy as moral reflection. Effective deliberation can occur without any questioning or assessment of the values utilized in the deliberation, but not for Ms. B. As part of her deliberation, Ms. B must formulate the values that have been emblematic of how she has lived and are important to how she continues to live and how she will die. Given her unanticipated and novel situation, this is more than an exercise in values clarification; rather, it is an exercise in the construction and evaluation of values as an integral part of the construction of a self. Ms. B will be testing and critically examining the values she entertains to see how well they fit her past life and her new circumstances. She will be using the resources of creative construction and comprehensive critical assessment, and those resources will reciprocally enhance one another. As Miller recognizes, this sort of deliberation unites the important but neglected senses of autonomy as authenticity and autonomy as moral

122 

Chapter 5

reflection: “Moral reflection can be related to authenticity by regarding the former as determining what sort of person one will be and in comparison to which one’s actions can be judged as authentic or inauthentic” (1981: 25). What Ms. B needs for her imposing task is not just moral insulation— independence and protection from interference in her decision making— but moral help. She needs help understanding why she unexpectedly has a problem, what her problem is, and how she can work through it. Helping her is, in part, a matter of conveying information, but it will not be enough to simply give her information and tell her to choose what she wants. Franz Ingelfinger, a longtime editor of the New England Journal of Medicine, once said, with vigor, “A physician who merely spreads an array of vendibles in front of the patient and then says, ‘Go ahead and choose, it’s your life,’ is guilty of shirking his duty, if not of malpractice” (1980: 1509). Ingelfinger stresses that a physician must take the responsibility of recommending a concrete, practicable but also morally insightful and satisfying course of action to a patient; but no such recommendation can be made to Ms. B without an understanding of the kind of person she is and wants to be in her death. The goal for all those helping Ms. B, but particularly for Dr. M, is to avoid a looming tragic outcome. Do moral and religious prohibitions of euthanasia prevent her from dying in the manner she desires? What, exactly, are the nature, the scope, and the force of those constraints? As the women who after genetic counseling are searching for a green light to try to conceive, so is Ms. B searching for a green light to avoid dying by suffocation. Dr. M’s responsibility, because he is the physician caring for her and because he is the one who ultimately will have to endorse or reject her decision, is to help her, and at times perhaps guide her, through her search. Both Ms. B and Dr. M will be active participants in that search and will be engaged in an on-going, reciprocal process of making judgments that are informed, shaped, and tested by the resources of nonformal reason. Dr. M will begin by making judgments about substance and process: about what is troubling Ms. B and, given that, about what is the best way to try to help her. Ms. B also will begin by making judgments about substance and process: about the relevance, plausibility, and importance of what Dr. M tells her; about the significance of the way she is reacting to what she learns from Dr. M; and about how she should go about making her decision. Their collaborative search requires a comprehensive critical assessment of the elements of Ms. B’s dilemma. That scrutiny should begin with a strong reaffirmation of the mission of palliative care: to identify and mitigate suffering of whatever kind. Palliative care is about facilitating a good

The Process of Ethical Resolution 

123

death, and systematic critical assessment can emphasize that that is what euthanasia literally means. In that general sense it is hard to imagine that anyone would not accept euthanasia. So Ms. B must fear that there is some specific subordinate issue that clashes with her religious values. That fear has to be investigated. Dr. M then can explain how means and ends are inseparably entwined in decisions about how someone will die and how the interdependence of means and ends can create confusion with respect to the vague and emotionally charged notion of euthanasia; and he can assure her that he will support her throughout her search for personal clarity on the issue and respect her eventual wishes. In that process, for instance, the notion of “playing God” will undoubtedly be subjected to systematic critical assessment. What is the proscription of “playing God” supposed to prohibit? A death that is not “natural”? If so, what does “natural” mean? Or just dying sooner than one would have without intervention? If so, what kind of intervention? What if both providing privacy for Ms. B in a single room and providing her with company in a multi-occupant room were to lead (from, respectively, loneliness and shame) to depression which hastened death; would either be an inappropriate intervention? Or does “playing God” mean exercising a kind of control over life that one should not exercise? If so, is the choice of a room different from the choice of sedation? Or perhaps “playing God” is having an intention or a desire that one should not have? Again, is the choice of a room different from the choice of sedation, and if so, why? And similarly for all the other possible kinds of intervention, control, or intention, why would they be supposed to be wrong? Ms. B’s ideals and the values that could serve as proxies for her ideals also must be explored critically. What does she want? Suppose Ms. B says she wants to be a moral person and a devout Christian and she wants a good death. Those are compelling ideals, but rather empty when taken in isolation. What specifically do they mean to Ms. B? A good death probably means dying peacefully, without suffering and with dignity, but what being a moral person and a devout Christian means is less clear. If the latter means not “playing God,” for example, we can appreciate the complexities in constructively specifying what Ms. B seeks. Values that serve as proxies for her ideals must be identified, and then developed and refined by determining how the alternative ways in which she could die do and do not instantiate them. Ms. B will make judgments about which values are more compelling to her, whether her values are mutually compatible, and where her values are leading her; about how well she is enduring the responsibility of having to decide; and about whether her deliberation is proceeding in a helpful,

124 

Chapter 5

productive manner. At the same time, her clinical options can be modified and adjusted in the effort to try to find a way of dying that comports with her crystallizing values. Perhaps a little creative construction might help her here: if an explicit protocol were developed that initially permitted her to set her own sedation level—to a level where she felt she was sufficiently comfortable—without the intention to relieve her pain by foreshortening her life, which would be against God’s will, and with the assurance that a level of sedation that would keep her sufficiently comfortable would be continued when she could no longer do so, might this arrangement be acceptable to Ms. B? It would, for example, satisfy the Catholic doctrine of double effect. There will be a complex, dynamic interplay between her values and her options, a process of reciprocal testing and customization aimed at designing a way of dying that satisfies Ms. B’s values and thereby preserves her ideals. The resources of comprehensive critical assessment and creative construction will direct this crucial component of the process. Throughout the entire process Dr. M will be observing Ms. B and making judgments about how she is responding, intellectually and emotionally, to information and to their interactions; how she is assimilating and assessing material and reasoning with it; and how she is coping with the burden of having to make a decision. Does she retain information? Is she thinking clearly? Are her emotions appropriate? Based on what he observes, he might need to revise his views about what is troubling her, what kind of further help she needs, and who best can provide that help. His observations and the inferences he draws from them should be compared with and checked against the observations and inferences of Ms. B’s family and the other health professionals who care for her. Based upon their observations of how she looks, how alert she is, what her moods are like, and how she is sleeping, for example, what is their sense of how well Ms. B understands her situation and is working through it, how she is feeling, how confident she is about what she wants, and what kind of help and support she needs? At the same time, Ms. B will be observing the behavior and demeanor of Dr. M. Is his communication of information candid and forthright, or does it seem hedged and evasive? Does he readily acknowledge weaknesses and uncertainties? How does he respond to requests for clarification and explanation and to being questioned about his judgments? Is he receptive and supportive, or does he become defensive? Ms. B will check the information he gives and the opinions he offers against what he has told her family and against the views of other health care professionals. In light of her observations and the evidence she obtains from her interactions with others, she will continually be making judgments about Dr. M’s forthrightness

The Process of Ethical Resolution 

125

and trustworthiness and the soundness of his opinions. Ms. B also will be observing the behavior and demeanor of her family. How are they coping with her plight? Do they look anxious and worried? Are they uncharacteristically silent or nervous around her? Are they edgy and short-tempered? Or, on the other side, are they infuriatingly passive and deferential? Ms. B will try to gauge how much of a toll her predicament is taking on them, and judgments about that, too, will figure in her deliberation. Formal and nonformal reasoning methods will pervade the process. Dr. M can question how and why Ms. B arrives at her views and whether the conclusions she draws from them really follow, and he can probe the consistency of her positions. He can use narrative and case-based reasoning and argument by analogy to examine the various methods of ending life that are difficult to distinguish and to relate them to Ms. B’s values. Dr. M can recount stories about patients in similar circumstances that reveal how and why they made their choices and what happened to them. He can describe scenarios that differentiate interventions used to instigate the dying process—for example, for patients who have unremitting, unbearable, all-consuming pain or patients for whom life has lost all meaning— from interventions used not to accelerate death but to change the nature of the dying process. Creative construction functions substantively in the operation of the process and procedurally in the design of the process. Dr. M has to be flexible and imaginative in responding to difficulties Ms. B might be having, reframing matters for her, devising new scenarios for her to consider, and altering old scenarios to illustrate relevant similarities and dissimilarities to her particular circumstances, and in entertaining novel clinical options for her (such as the sedative delivery protocol suggested above). Ms. B has to be flexible and imaginative in filling in her ideals of a moral person, a devout Christian, and a good death, and in engaging the values that come to serve as proxies for these ideals with the cases, stories, and possibilities that Dr. M provides. All the substantive work occurs within a flexible, adjustable process that Dr. M devises and Ms. B accepts, and that both Dr. M and Ms. B continually appraise and modify to improve it. For their work to be productive and helpful, the process must be collective and consultative. Dr. M should involve Ms. B’s family so that their fears too can be addressed and so that they can provide sympathetic support for, and a trustworthy check on, Ms. B’s deliberation and judgment. Dr. M should also enlist the help of a religious advisor who can guide Ms. B in developing theologically sound and personally meaningful proxies for the ideal of being a devout Christian and

126 

Chapter 5

in relating her clinical options to the doctrines and values of her faith. A clinical ethicist not hamstrung by the model of applied ethics can do the same on the secular side. And Dr. M should involve experienced palliative care physicians in his exploration of clinical alternatives that manifest, as much as possible, the values of Ms. B. This collaborative, critically attuned process will continue until Ms. B and Dr. M are able to make judgments about when and how it can end—until Dr. M is able to make a recommendation and Ms. B is able to make a decision. Ms. B’s deliberation is much more complex than is recognized by the conventional account of respecting patient autonomy (and the accompanying legal account of obtaining informed consent). It is not, as that orthodoxy implies, simply an individual matter; it is a social endeavor that can be rationally designed to enhance the rationality of the judgments Ms. B makes and the decision that emanates from those judgments. The process is collective for two reasons. One is that, as we have seen in science, the resources of nonformal reason are more effective and more powerful when they operate socially. With respect to observation, the more different modes of access to how Ms. B is thinking and feeling there are, and the more the outcomes of that multiple access can be comparatively assessed, the better can be the judgments of others about how she is doing and what she needs. Likewise with respect to critical assessment and methods of reasoning. The more diverse resources and perspectives that can be brought to bear, in a coordinated, focused manner, on understanding her problem and ways of dealing with it, the more confident one can be about the rationality of the process and thus the outcomes of that process. The other reason is that, contrary to what the conventional understanding of autonomy suggests, Ms. B’s values are not individual and private. Ms. B did not create her ideals and values de novo. Her ideals and values have emerged within multiple historical and social contexts, most notably and influentially those of her family and her religion, and have been tested in countless social interactions and adapted to a multitude of social circumstances throughout her life. Because values are socially created and grounded, stories about and illustrations of their genesis—nonformal reasoning techniques—are central to understanding and assessing their legitimacy. In addition, given the social genesis, meaning, and operation of values, it is eminently reasonable for Ms. B, as did the women deciding whether to try to conceive after genetic counseling, to search for the social boundaries within which she may decide. That search should not be misconstrued as an abandonment of her autonomy; rather, constructing a realistic strategy for framing and managing her problem should be embraced as

The Process of Ethical Resolution 

127

an expression of her autonomy. The stories and perspectives that Dr. M and others provide can enable Ms. B to formulate and assess her values and to appreciate the limits within which her values will operate. The point of this posited depiction of Ms. B’s deliberation and the insistence on its inherently social nature is not to denigrate or disparage autonomy—autonomy is an important moral ideal. The point is to comprehend the complexity of autonomy. Like the women deciding whether to try to conceive after receiving genetic counseling, Ms. B’s predicament is profoundly and momentously individual. And like the women deciding whether to try to conceive after receiving genetic counseling, Ms. B’s deliberation is robustly social. The challenge is not just to respect her right to decide but to help her make the right decision—the decision that is right for her—by deciding rightly. Deciding rightly is deciding rationally. The entire process that Ms. B and Dr. M develop is a creative construction designed to resolve an initially ill-defined problem in a manner appropriate to the issues and the persons involved. This accomplishment is at least one of the most important exercises of rational intelligence in the moral domain, if not the most important. It necessarily underlies all that transpires at more practically accessible levels of moral discussion and action. Formal reason lacks that capacity. Nonformal reason has that capacity and also illuminates and rationalizes it. Ms. B’s struggle might not end well. Rationality, as Bertrand Russell reminded us, is about thinking and acting intelligently; it does not, and cannot, guarantee nice outcomes (1912/2005, chapter 6). At the end of her deliberation, Ms. B’s situationally tempered values could be incompatible (as in the example of selecting a room for Ms. B above), and she could face a no-win choice. In addition to her fear of suffocation, she now fears that using the means necessary to avoid suffocation would violate her deep moral and religious convictions, which represent what it means for her to be a moral person and a devout Christian. After careful scrutiny and assessment of her situation, one of those fears might persist. If so, she will have to choose between not dying by suffocation but doing something, just as she is dying, that she regards as fundamentally wrong, and abiding by the precepts of her morality and her faith but enduring the kind of death she fears so profoundly. In that event, no matter what Ms. B chooses, she will lose something vital to her conception of what it means to be a good person and to lead a good life. But a process of deliberation shaped, tested, and conducted by the resources of nonformal reason gives her both an examined death, as an expression of her self, and a better prospect of a life that ends as well as it can.

128 

Chapter 5

5.2.2  Ms. F The preceding vignette about Ms. B shows how the meaning of a core moral term—autonomy—is ambiguous and cannot be meaningfully applied until it is contextually disambiguated. In vignette 2.2 Ms. F has been offered the option of a third lung transplant, and she has to use the resources of nonformal reason to make an autonomous choice in circumstances riven with uncertainty. Refusing the transplant means death. Accepting the transplant offers the possibility of continued life, but the probability of a successful transplant is uncertain, if the transplant is successful, how much longer she might live is uncertain, and what that longer life would be like is uncertain. In general, what would she have to sacrifice and endure to have a life of unknown quality for an unknown time? Overlying those issues are the personal and social considerations. Pressures from her family and those close to her push in both directions, so no matter what she decides, she will disappoint and dismay people who are important to her. Moreover, if she decides to try another transplant, she has to move to a different hospital and lose the transplant team that has become a “family” to her. Yet how can she refuse? How can a person facing imminent death choose to die when given an alternative, even if the procedure is risky and the benefit is speculative? The clinical ethicist who worked with Ms. F was appalled that such a choice had been imposed on her: “I was outraged with the physician who had made what I thought was an outlandish and exploitive, if not coercive, offer of a third lung transplant” (Marshall 2001: 147). How might the clinical ethicist help Ms. F? Her predicament is not amenable to the decision making of applied ethics because it is not a matter of applying or adapting a consistent set of principles to a problem, nor is it a matter of maximizing utility on the basis of specified risks and values, none of which are even well defined here. Nor is it just a matter of determining whether Ms. F’s decision, whatever it turns out to be, is autonomous in the spare sense of being free and informed. What the clinical ethicist can do is encourage her to assess critically how she is framing her choice and her process for making a choice, and that examination could be instigated by focusing on her analogy between a broken-down bicycle and her brokendown body. Why does that mechanical conception of her body appeal to her? Is a person simply an assemblage of material parts and a transplant surgeon just a body mechanic? Does she regard that metaphor as a realistic way of thinking about her dilemma, or does it represent and confirm what she already knows she has to do? It always is hard to refuse a gift, perhaps unthinkably so when it is the gift of life. If so, Ms. F is entrapped. But if she is unsure, perhaps all she can do is defer agreeing to the transplant and

The Process of Ethical Resolution 

129

hope that something will change in a way that frees her, either by clarifying the decision or by preempting it. Making that decision, and accepting responsibility for that decision, could be too hard. Comprehensive critical assessment might dislodge her resignation—the perception that she has no choice. If so, the clinical ethicist could suggest that Ms. F imagine different scenarios of what her life might be like after a transplant and assess her reactions to those scenarios, including their effects on her relationships with the people who are close and important to her, a project for which she has abundant experience. For that enterprise to be productive, though, the conflicting personal and social pressures imposed on Ms. F have to be removed or at least eased. Doing that requires designing a process that enables the two sides to understand that they are exacerbating Ms. F’s dilemma and encourages them instead to explore ways in which they can support her through her ordeal. The clinical ethicist could enlist a social worker or a member of Ms. F’s health care team to facilitate separate meetings with Ms. F and members of each side to explore the complexity and the difficulty of her situation and the role they play in it. Ms. F’s decision is daunting because she does not want the control and the responsibility the decision gives her. Requesting to die when one is suffering or life has become meaningless is hard, but refusing a chance to live is even harder. Perhaps Ms. F needs permission to refuse the offer of a third transplant. Nobody, including the clinical ethicist, can give her that permission. What the clinical ethicist might be able to do, though, is design and guide a process of rational deliberation that could allow Ms. F to give herself the permission she is seeking. However it all turns out for Ms. F, a deeper and wider process of deliberation, like the one used by Ms. B, needs to be constructed for Ms. F. Only nonformal reason’s rich sense of design for flourishing—even in the most straitened circumstances—can create such a process that fits Ms. F’s capacities and her situation. The broader moral issue that Ms. F’s struggle raises is prevention. Wherever possible, moral problems should be prevented, not just solved. What aspects of the institutional and structural design of a health care system made the offer of a third lung transplant to Ms. F possible? Speculating about mercenary or self-aggrandizing motives on the part of the transplant surgeon would be distracting and would evade the deeper institutional design issues that underlie Ms. F’s problem. What institutional roles and relationships allowed the departing physician to act unilaterally? Why did the offer not have to be vetted and approved by the team that had cared for Ms. F for so long? What professional and economic incentives drive the

130 

Chapter 5

establishment of new transplant programs and spur recruitment of patients for those programs? And what social and cultural ethos underpins the health care system and its constitutive institutions and practices? As Ms. F’s clinical ethicist observes, “This situation could have arisen only in a clinical setting in which the biomedical resources, technology, and professional capability support the ideology—to a certain extent, even the mythology— that human bodies can live indefinitely” (Marshall 2001: 147). The individualistic orientation of applied bioethics neglects background ethical issues and, in any event, does not have the resources to address them. As we will see in part II of the book, nonformal reason promotes a more expansive conception of morality that takes ethics to be fundamentally a matter of design, of public as well as private matters. 5.2.3  K’aila The story of K’aila (vignette 2.3) vividly demonstrates how moral depth or insight transcends the conflicts among moral principles. It is unclear what course of action is in K’aila’s best interest, because there are medical rationales both for trying a transplant that could extend K’aila’s life, albeit only modestly while also risking losing it, and for providing compassionate care to K’aila as he dies. K’aila’s parents must make a judgment not only about what is best for K’aila but about what course of action they can live with, that is, how they can preserve the integrity of the persons they are. Momentous decisions in life—for example, choosing a career, getting married, and having children—are ultimately decisions about the kind of person one wants to be. Considering their integrity does not mean, however, that the decision of K’aila’s parents will be egoistic or selfish; to the contrary, their concern for the welfare of K’aila is even more paramount because K’aila is an intrinsic and important part not just of their lives but of who they are. Individually and socially organized processes for the regulation of moral judgment enable K’aila’s parents to deliberate rationally throughout their struggle. The individually organized processes take several forms. The reasoning that K’aila’s mother describes occurs at multiple levels, for instance, substantively at the levels of K’aila, themselves, and society and procedurally at the levels of deliberating and deliberating about how they are deliberating. Those levels operate reciprocally and reveal the nature of their personal integrity to be the integrity of integration and coherence, as K’aila’s mother recognizes when she comments, I “could not deny what I knew at other levels.” K’aila’s mother reports that her reasoning was checked by her feelings; for example, she “felt lucid,” an observation that shows that the kind

The Process of Ethical Resolution 

131

of integrity being sought is the integrity of an entire human being. And her butterfly metaphor is both the culmination of her reflection and reasoning and an affirmation that she has arrived at a decision with which she can live and for which she can take responsibility. Several socially organized processes provide checks on the individual reasoning of K’aila’s parents, and these processes operate within various institutional mechanisms. One check comes from the pediatrician who cared for K’aila, a responsibility that permitted him to form his own view about what was in K’aila’s best interest and to pursue his view when K’aila’s parents chose differently. Another comes from the child welfare agency that became involved at the instigation of the pediatrician and that performed its role as an agent of the state in protecting children. Then there are the requirements and procedures of courts in resolving disputes between individuals and the state about the welfare of children, and the individual judges who hear evidence about, reason through, and resolve such disputes. In their deliberations K’aila’s parents had to take account of and respond to the demands and constraints of these socially organized processes for regulating individual judgment. The resources of nonformal reason are utilized in the conscientious efforts of K’aila’s parents to decide what to do for K’aila. They mostly use informal reasoning methods, which are subsequently integrated with formal reasoning methods when the judge invokes social and legal norms to arrive at a legal ruling. Creative construction is exemplified in the butterfly metaphor as a way of focusing the moral issue for them as they search for moral coherence and integrity in their lives in the face of weak and inadequate social guidance. Their search for coherence across the types and levels of their reasoning, singly and jointly, displays comprehensive critical assessment. In addition, the activities of the pediatrician, the child welfare agency, and the judge provide external social checks on their deliberation. The upshot is a rational process of individual decision making that is embedded within and constrained by a rational, institutionalized process of social decision making. 5.2.4  Mrs. Smith When, for many reasons, the ethical considerations relevant to a situation do not provide adequate guidance about what ought to be done, in practice the focus shifts from making the right decision to making a decision in the right way. The substantive question of what ought to be done is replaced by the procedural questions of who ought to decide and how they should go about deciding. That is what happened in K’aila’s case, and that also

132 

Chapter 5

happens in Mrs. Smith’s predicament (vignette 2.4), but in an instructively different way. The challenge of deciding whether to insert a feeding tube into Mrs. Smith’s stomach displays the indeterminacy of the principles formal reason brings to this issue: the substituted judgment principle and the best interest principle. Mrs. Smith’s situation has a number of similarities to K’aila’s. In each, a life-and-death decision is being made for someone who is incapable of such a decision. In both, the problem is resolved by shifting from substance to process. And in both, the potential for supervision and intervention by health care professionals and courts imposes robust socially organized constraints on individual decision making. Yet there is an important dissimilarity between the two situations. With respect to Mrs. Smith the core values that would make the decision clearly moral are hard to discern. Mrs. Smith is not suffering. Mrs. Smith can no longer interact with other people or the world. She has lost the capacities that are constitutive not just of autonomy but of agency itself. There is no identifiable family ethos, tradition, or set of values, fidelity to which would provide a way of formulating and understanding the importance of what is at stake and would give direction for making a decision. Nor does vitalism—the concern to preserve life because it is life—seem to be operating here. In the probable absence of such core values, the considerations that remain appear to concern pragmatic matters and the interests of family members. To be sure, just as K’aila is an intrinsic, important part of the lives of his parents, Mrs. Smith is an intrinsic, important part of the lives of her children. But whereas for K’aila there are profound questions about the means and the prospects of keeping him alive and even more profound questions about what it would be like for him to be alive, for Mrs. Smith there are neither. And whereas for K’aila’s parents there are profound questions about how their decision can be made in a way that preserves the coherence and integrity of their own lives, for Mrs. Smith’s children there is no evidence of a comparable challenge to them individually or to their filial harmony. Perhaps, then, Mrs. Smith’s predicament does not raise a moral problem. And perhaps that is why it falls in what Arras (section 2.3) calls the gray area. Given those telling differences, an individual moral decision about whether to place a feeding tube for Mrs. Smith might not be amenable to either formal or nonformal reason. What is available only to the latter, however, is making the judgment that there is no moral issue at stake. The only further way nonformal reason might operate here is through the

The Process of Ethical Resolution 

133

socially organized regulation of judgment, provided by health care professionals and a court, in the event that the family’s decision making transgresses established limits of social and legal acceptability. Nonformal reason can fail to provide adequate guidance for moral problems, but that happens less quickly and less frequently than with formal reason because the meager resources of formal reason have two endemic, debilitating problems: the vagueness and ambiguity of its principles and conflicts among its principles. Either there are no higher-order principles from which solutions to these two problems can be deduced, or, to cover all the conflicts, there must be multiple higher-order principles that, a few moments reflection reveals, will also have conflicts and, in turn, need higher-higher-order principles, and ad infinitum. Similarly, the shift from substance to process will remain a mystery to formal reason. Nonformal reason, in contrast, holds out the possibility of offering good, rational counsel for moral problems because properly designed process is central to all moral problem solving, not least when it affects human flourishing, as these vignettes convey. 5.3  Moral Effectiveness: Moving beyond Principles How can bioethics be effective in handling moral problems? Reports of effective problem-solving from clinical ethicists do not involve the application of moral principles or moral theory.7 Early in his bioethics career Art Caplan, for example, had two acknowledged successes. One occurred while he was teaching an elective course called “Ethics Rounds” with a psychiatrist and internist in a large urban medical center. The students and instructors would visit patients selected by the students, interview them, and then discuss the moral issues the students and instructors discerned. Once a 90-year-old woman who had no relatives was picked by the students because they were worried about what would happen to her after she was discharged. Caplan dutifully brushed up on his moral philosophy: “I came to the class fully prepared to discourse on theories of distributive justice at a moment’s notice, since I rather naively believed the students might benefit greatly from a disquisition on Mill, Rawls, and Nozick in trying to figure out what to do with the old woman” (Caplan 1983: 311). But he quickly got sidetracked: As soon as the medical instructors and students had gathered together, we hurriedly set off to find and interview the old woman. We all burst into her room just as she was in the process of defecating. To my surprise, no one was deterred by her behavior, and both the psychiatrist and the internist proceeded immediately

134 

Chapter 5

to interview the woman about her life plans, goals, and personal aspirations. I remained uncharacteristically silent during this exchange, and it was only when the class had returned to the confines of the psychiatry lounge to discuss the case that I proffered the opinion that it might have been better to wait until the woman had finished her excretory functions before interviewing her. (311–312)

His reaction “was greeted with some consternation by both students and the other teachers,” but they conceded that the privacy of patients is important and should not be overridden by their busy schedules. There was likely more than this at issue (see below), but the value of “applied” bioethics, as Caplan had practiced it, was established: “my esteem among the members of the course was assured for the duration of that particular clinical rotation” (312). Caplan topped that success when he tackled the thorny problem of rationing scarce medical resources. Hot summer weather made it difficult for persons with emphysema and other respiratory ailments to breathe and drove them to the emergency room for oxygen. But there were only two oxygen units, and throughout the day and night at least a dozen people often were waiting to use them. The staff in the emergency room wanted Caplan to help them develop criteria for fairly allocating their scarce resource. Again Caplan prepared as an applied ethicist would, by consulting the philosophical and theological literatures on justice and microallocation. He found predictable defenses of allocating scarce resources based on criteria of need, merit, and social utility and of deciding randomly by using a lottery. But he was savvy enough to realize that the medical staff could get that far without either him or the materials he read. Then he had an inspiration: It occurred to me that it might be possible to solve the allocation problem by ameliorating the source of the scarcity. I asked some of the emergency room staff if Medicaid/Medicare covered the provision of air conditioners in the homes of persons suffering from respiratory ailments. It turned out, much to everyone’s surprise, that the machines could be prescribed and the cost reimbursed. (1983: 312)

Caplan’s reputation was secured: “I ascended to the status of moral guru in the emergency room, famous as the man who had solved the oxygen machine crunch” (312). Neither success resulted from applied ethics, as Caplan recognizes: “… these two cases of efficacious moral action were hardly dependent on analytical rigor or theoretical moral sophistication for their genesis. Indeed, in both cases, ethical theory would have been the wrong place to turn for a solution to the issues under consideration” (312). What Caplan means

The Process of Ethical Resolution 

135

by “ethical theory” must be understood carefully, however. The ethical theory he has in mind is the norm-constituted brand of ethical theory propounded in philosophy and invoked by applied ethics. Our goal is not to reject ethical theorizing or ethical theory but to develop a broader theoretical understanding of ethics that can account for the kinds of success that Caplan had. Toward that end, the sources of Caplan’s successes should be appreciated. Caplan’s first success was the product of observation and critical assessment. In a rudimentary sense, when the students and the instructors entered the patient’s room, Caplan saw what everybody else saw. In a more sophisticated sense, Caplan saw what nobody else saw, just as a chess grandmaster can look at the position on the board and assess the outcome, and as an experienced firefighter can look at a burning building and assess the danger. What Caplan observed and the other participants did not was the moral salience of the situation, probably because of his training and experience and because, as an outsider, he was not inured to routine encroachments on patients’ privacy, particularly when they are egregious. Moreover, Caplan was willing to expose and challenge the deep-seated assumption that the time of busy doctors, residents, interns, and medical students is more important than the dignity of patients. The consternation he caused suggests that behind the issue laid the deeper problem of the self-interest inherent in treating this elderly woman, or any other patient, as simply a source of information and learning rather than a person in her own right. This prospect speaks to the cultural ethos that governs a hospital, or any other institution, and suggests an inadequacy that is even more deeply embedded and needs to be confronted. Caplan’s second success was the result of creative construction and formal and informal reasoning. Rather than accepting the ineluctability of the scarce resource problem and searching for a way of solving it, he entertained the possibility of dissolving the problem by removing its cause. That is, he stepped outside the confines of the problem as posed and replaced it with a wider problem that admitted a radically different solution, an exemplar of intelligent procedure and a creative act characteristic of nonformal reason. And he proposed that the terms of Medicare and Medicaid reimbursement might be applied directly or extended by analogy to cover the cost of air conditioners for these patients. The sources of Caplan’s successes—observation, critical assessment, formal and informal reasoning methods, and creative construction—are the resources of the broader conception of nonformal reason that accounts for the rationality of both science and ethics. And in both cases they led to ethically valuable outcomes that would

136 

Chapter 5

have been entirely missed had his methods been labeled as irrelevant or nonethical. 5.4  Conclusion The resources of nonformal reason reveal and account for the rationality of the dying children who imperatively are trying to find out what is wrong with them and what will happen to them. Bringing those resources to the vignettes reveals the limitations of formalist rationality and the notion of autonomy in applied bioethics, limitations that prevent it from providing moral help when moral help is most needed. A formalist account of autonomy as self-determination is acontextual because the locus of autonomy does not matter. Whether it is the autonomy of the individual, the autonomy of the family, or the autonomy of the tribe, the function of autonomy as self-determination is to protect the exercise of autonomy from interference. Moreover, a formalist account of autonomy is empty because it prescinds from the values that are necessary for and relevant to a decision: autonomy as self-determination is in the service of the specific values operative in a context. Formalism does not provide resources for critically assessing and implementing existing values or for formulating and adopting new values. Yet those tasks are precisely what makes the protection of autonomy so essential. Nonformal reason provides the capacity and the resources needed for rationally accomplishing those vital tasks.

II  Ethics as Design

C

T

6  Ethics as Design and Its Two Distinctive Methods

C

Re-reasoning ethics begins with an enriched conception of rationality that fits our whole intelligence, not just our logic, and develops a correlatively more encompassing conception of ethics that fits our whole lives. Here we complete our account of how this new rationality re-reasons ethics by presenting Caroline Whitbeck’s (1996, 2011) explanation of how design in engineering can be brought to practical problems in ethics and by introducing two distinctive, fundamental methods for rationally designing resolutions of practical problems in ethics: constructing a fully engaged moral compromise and creating a liberated, extended, and enriched wide reflective equilibrium. Design is used in engineering to work through a practical problem in a practical way. Doing that requires framing a problem, identifying possible solutions, and so on, and improving these across trials, all in the manner discussed in section 4.3. Design in engineering is not just doing that, but figuring out how to do that, and within the constraints applying such as for time and other resources. Doing all that for practical problems in ethics is, as Whitbeck puts it, “doing justice to moral problems” (1996: 9). Two distinctive methods of problem solving help to do justice to moral problems: compromise and wide reflective equilibrium. Avishai Margalit (2010) recognizes that “compromises are vital for social life.” Compromise is widely accepted and utilized in bargaining and negotiation about economic and other matters and in politics, which Margalit calls “macromorality.” Compromise also is “center stage in micromorality,” the morality of persons and of dealings among individuals: We very rarely attain what is first on our list of priorities, either as individuals or as collectives. We are forced by circumstances to settle for much less than what we aspire to. We compromise. We should, I believe, be judged by our compromises more than by our ideals and norms. Ideals may tell us something important about what we would like to be. But compromises tell us who we are. (2010: 5)

E

140 

Chapter 6

Compromise is ubiquitous throughout our moral lives, indeed unavoidable for all finite creatures, yet it can have no place in the logic-based, formalist approach to ethics. The method of fully engaged moral compromise acknowledges the unmet demands of the conflicting ethical claims that are compromised. Moral compromise has to be characterized and clarified, and its central role in ethics defended. The development of the concepts of narrow reflective equilibrium and then wide reflective equilibrium liberates moral philosophy from the debilitating limitations of formal reason’s foundationalism. That was an important advance, but more still needs to be done to free reflective equilibrium from formal reason’s perdurable insistence on justification and theory construction. Once fully liberated, wide reflective equilibrium has to be extended to encompass diverse forms of moral deliberation and creative moral options, to accommodate a vast expansion of morally relevant considerations, and to be enriched with compromise as a method for reaching equilibria. The result is a more realistic, more powerful, and rational nonformal ethics. 6.1  Whitbeck on Ethics as Design Whitbeck defines design problems broadly, as “problems of making (or repairing) things and processes to satisfy wants and needs.” A chef designs a holiday meal; a boss designs a work schedule; a scientist designs an experiment. In an engineering school, design courses are taught along with theory courses. Engineering theory courses, Whitbeck explains, “teach students to understand theory and how to apply it to solving problems with mathematically exact and usually uniquely correct solutions.” Design courses use “synthetic” as opposed to “analytic” reasoning “to construct good responses to practical problems” (Whitbeck 1996: 10). What aspects of engineering design, and design in general, are relevant to ethics? Whitbeck describes four and illustrates them with the example of two engineering students assigned the task of designing a child seat that can fit on the top of suitcases being wheeled to an airport gate, can be used on a seat in the airplane, and can be stored either under a seat or in an overhead bin. The first feature is that only rarely, if ever, is there a uniquely correct solution or even a predetermined number of correct solutions. In engineering, Whitbeck reports, “if there is one solution to a design problem, there are usually several” (1996: 10). And both engineering and moral problems can be made simpler, even trivial, when the framing or specification of the

Ethics as Design and Its Two Distinctive Methods 

141

problem is so tight that the range of acceptable solutions is small. If we add “must be metal and light” to the child seat requirements, we essentially limit the framing solutions to aluminum plate and tubing. If we want to prescribe what bystanders to an accident should do, the moral options are nicely reduced if we specify that bystanders are capable of making a difference and will not be in danger and that there are no other factors. The moral problem then becomes trivial. Trivial problems selectively reinforce the philosophical penchant for portraying ethics in terms of “doing the right thing,” but when there are multiple acceptable solutions, there is no single “right thing to do.” Along with providing medical aid, a bystander might, for example, clean up to prevent a fire, warn approaching cars to avoid collisions, phone for an ambulance, and the like, and each might represent a crucial, but competing, constructive intervention. The second feature is that even if there is not a uniquely correct solution, there might be clearly wrong solutions, and among the solutions that are not clearly wrong, some will be better than others. That recognition is a corrective to the glib adoption of the dichotomy that either there is an objectively correct solution or all solutions are arbitrary and subjective, a common presumption promoted by the logical-deduction-from-a prioriprinciples approach to ethics. The third feature is that when there are two (or more) acceptable solutions, they likely have different advantages and disadvantages and so there is no one demonstrably better (or best) solution. But the differing advantages and disadvantages of multiple solutions might differently fit the different situations of different persons. (That one solution is best for one person while another solution is best for a different person is contextual relativity, not relativism, because the fit of solution to person can itself be critically appraised. The opposition between either acontextual/universal or arbitrary is another false dichotomy encouraged by the logical-deduction-from-a priori-principles approach to ethics.) The fourth feature of design in both engineering and ethics is that solutions must: (a) achieve the desired performance or goal; (b) conform to prescribed specifications or criteria; (c) be reasonably secure against accidents or other untoward events that could produce bad consequences; and (d) be compatible with existing background constraints. These aspects of design are ignored when moral decision making is taken to be the application of a moral theory or principle. What lessons for design in ethics follow from this analogy with design in engineering? The first and most crucial lesson is the need to appreciate the rampant ambiguities and uncertainties that are inherent in framing

142 

Chapter 6

a problem. The engineering students designing a child seat, for example, did not know what features potential users might want or what solutions might already exist, nor did they know how compatible the various requirements for the child seat might be, for instance, whether such a device could accommodate a full range of sizes of preschool children yet be light and easily removable. Problems have to be defined in a manner that preserves as much flexibility as possible. And because often it is not possible to wait for the resolution of ambiguity and uncertainty, fallible but rational judgment is essential. A second lesson from engineering design is that an initial characterization of a problem is only a first step, beyond which lies the devising of possible solutions. As Whitbeck points out: [Identifying solutions] is one of many features that distinguish moral problems from formal “decision problems.” Decision problems or problems in decision analysis include specifying the alternatives among which one is to decide; that is, a fully defined decision problem is a type of multiple-choice problem. This need to develop possible solutions implies … that statements of moral problems as open-ended situations do more justice to those problems than do statements of them as multiple-choices. (1996: 13)

Moreover, solutions for the formal decision problems of applied ethics are often binary, for example, dispose of a regulated toxic substance by pouring it down the drain or do not pour it down the drain. But this ignores what the sequelae of these stark options might be. With open-ended, that is, deep problems, framing the problem and identifying potential solutions are interactive, complementary tasks. Whitbeck appreciates that, in order for solutions to be proposed, problems have to be formulated and that “even open-ended statements are only outlines of ethical problems” (1996: 13). Framing real moral problems requires investigating the features and details of a problem and the various contexts within which the problem arises. Only then can promising solutions emerge and be modified or altered as dimensions of the problem shift. With respect to a hypothetical scenario in which her supervisor directs her to dispose of a regulated toxic substance by pouring it down the drain, Whitbeck explains that if that were a real problem, “there would be a particular person who would be my supervisor whose character I might learn more about” as well as “an actual organization (a company, a university) with particular policies that I could investigate” (13). Not to mention investigating other ways of disposing of toxic substances, or of redesigning the experiment so as to avoid their use, and so on. Whitbeck praises the practice of “brainstorming” because it enables engineers to learn how to develop responses to problems, which is essential to

Ethics as Design and Its Two Distinctive Methods 

143

creative construction. But brainstorming, she cautions, “requires an uncritical atmosphere in which people can present half-baked ideas that may later be refined or combined” (Whitbeck 1996: 13). Whitbeck knows, however, that the attempts of moral philosophy students “to probe the complexity of the case are cut off” routinely and peremptorily and that brainstorming is strongly disapproved: “Articulation of any half-baked ideas is discouraged in the many ethics classes where adversarial debate is the primary method used. Although an adversarial format may provide some useful pre-law training, it does not help develop the ability to think constructively about resolving moral problems” (13–14). Philosophy students are expected to choose between the prescribed options (typically, utilitarianism and Kantianism), to apply the moral theory they deem to be correct, and to defend their position zealously and uncompromisingly. That approach might indeed teach the skills of debating or advocacy, two enterprises where the goal is to win, but it does not teach problem-solving skills, in morality or anywhere else. Whitbeck provides an example of what she regards as a “rather heroic capacity to brainstorm in the face of criticism” displayed by Amy (Whitbeck 1996: 14), a student who participated in Carol Gilligan’s research on moral development (1982). Amy responded to Lawrence Kohlberg’s classic Heinz dilemma of whether Heinz should steal a drug that could save the life of his wife but that he cannot afford. Amy, entirely reasonably, recognizes that Heinz’s wife should not die if her life can be saved and that Heinz should not steal the drug, and she realizes that if Heinz does steal the drug, he might go to jail and then would be unable to support his wife as she becomes sicker. So, quite reasonably, she explores options such as trying to negotiate with the druggist, borrowing money, and taking out a loan. Amy’s creativity is suppressed in a theory-dominated philosophy classroom, and Amy’s kinds of options are not available on the multiplechoice questionnaires used in psychology research to determine at which of Kohlberg’s Piagetian-inspired six hierarchical stages of moral reasoning a subject is.1 In short, with open or deep problems, as Whitbeck explains, framing a problem and devising solutions for a problem are interactive, reciprocal endeavors, not separable, independent activities. Each can inform and direct the other to enhance the understanding of a problem and how best to respond to it. And each constrains the other: conceptions of the problem limit the range of acceptable solutions, and the range of potential solutions limits how the problem can be conceived. The two complementary undertakings have to be integrated in a creative, flexible process of

144 

Chapter 6

problem formulation and solution. While all of Whitbeck’s lessons apply to all design processes, the design literature beyond engineering almost universally highlights the creative exploration of possibilities, both in framing a problem and in identifying potential solutions as the key feature of problem-solving processes.2 The third lesson is the need to act under time pressure. When time is short, it is important to pursue several possible solutions simultaneously so that alternatives exist if one turns out to be impossible or impracticable. Amy’s rejection of the dichotomous choice that is imposed—steal or don’t steal—and her ensuing brainstorming produce several options for Heinz to pursue, with various probabilities of success and various timescales for exploring each. The fourth lesson is that problems are dynamic. The nature of a problem can change as circumstances and contexts change, as constraints are investigated and modified or eliminated or new constraints are discovered, as elements of the problem are changed or removed or additional features are introduced, and as progress is made by addressing some aspects of the problem. And such changes can also change the conceptions of what a good solution might look like and what methods might be appropriate to uncovering solution(s). All these elements can coevolve as potential whole or partial solutions are explored. This dynamic fluidity is, in fact, the general nature of design process. Moral decision making should not be portrayed as a matter of solving static, completely but narrowly defined moral problems by choosing one of a limited number of static, completely but narrowly defined options. A moral problem, as Whitbeck trenchantly puts it, is not a “math problem with human beings” (Whitbeck 1996: 15). Michael van Manen and Nicole Kain (2017) provide a gripping example of how ethics as design can work. It concerns the creation of a policy that would permit the donation of organs for transplantation, particularly hearts, from infants who die in a neonatal intensive care unit (NICU). As an illustration of the staff’s commitment to improving care, van Manen and Kain describe a previous redesigning of the NICU. Its open-ward structure was altered to create single-family rooms that would foster family-centered care and provide comfort for both the infants and their parents and other family members. When their families were not in the NICU, though, the babies were often isolated in the single-family rooms. So the NICU was further redesigned—the walls between adjacent rooms were made collapsible so that staff members could interact with infants when their parents were absent. Both material redesigns of the NICU help to engage and assuage the infants and their families. Similarly, designing and implementing a policy

Ethics as Design and Its Two Distinctive Methods 

145

that would allow organs from infants who die in a NICU to be donated for transplantation would prolong the lives of other infants who are in a NICU. Van Manen and Kain appreciate that a policy has to “reflect thoughtful intent, careful deliberation” (2017: 20). For that outcome to be attained, the process of developing the policy has to be rigorous and transparent, and that means the process has to be broadly participatory and iterative. The creative design of their process engaged the three activities of “exploring, ideating, and testing carried out in iterative fashion” (23; emphasis in original). The goals of exploring are to “expand upon a specific problem and existent solutions,” to “find patterns and exceptions,” and to “identify limitations and boundaries” (23–24). The goals of ideating are to “redefine problems, creating space for novel standpoints, alongside practices such as sketching, drafting, and modeling imagined possibilities for design.” Testing is experimenting by addressing “instrumental, technical, aesthetic, and ethical aspects of design in the form of prototyping and implementation.” Testing requires that the process be robustly iterative: “design inherently necessitates re-design because a solution for a given problem may actually change the nature of a design problem itself” (van Manen and Kain 2017: 24). The work of developing the policy was done by an advisory group and the research staff, comprised of van Manen and Kain—a neonatologist with experience in qualitative research and a research assistant with experience in nursing and public health. The research staff supported the advisory group, which was comprised of 18 members who were “expert” in the care of NICU patient families and had diverse backgrounds, experiences, and views about donation after circulatory death (DCD). The advisory group adopted a focus group method because the “potential ‘novelty’ that emerges from focus group discussions is congruent with the creative crux of design” (van Manen and Kain 2017: 24). The advisory group held nine audio-recorded meetings that examined issues about “the processes surrounding DCD: eligibility criteria for DCD, consent process for DCD, pre-mortem interventions for the DCD donor, palliative care for the DCD donor-family, declaring death of the DCD donor, and post-mortem care for the DCD donor-family” (25). The content from the meetings was turned into four Survey-Briefings that were presented to “neonatologists, nurse practitioners, bedside and specialized nurses, neonatal subspecialty trainees, respiratory therapists, dieticians, pharmacists, and social workers” and circulated to parents who had had a child in the NICU. Subsequently a DCD policy was drafted and then “submitted to hospital administration, legal representatives, and other external experts and end-users” for their assessments. (29). Their reactions are likely to determine whether a policy

146 

Chapter 6

is implemented. Regardless of the outcome, the depth and the breadth, the content and the design of this policy-making endeavor render it a model of nonformal reason. (See also the NICE procedure, discussed in chapter 9.) Rather than being depicted as conflicts, moral problems are problems that are multiply constrained. Comprehensive critical assessment is needed to determine which constraints might be satisfied simultaneously and to what extent each might be satisfied, and which constraints are impossible or unlikely to be satisfied. Creative construction, operating in tandem with comprehensive critical assessment, will search for innovative ways of satisfying various combinations of the constraints to varying degrees. This characterization already suggests a potentially rich space of alternatives for searching. And even when attractive solutions might exist, there could be ways of improving them. What is essential for solving problems in ethics, whether individual or social, is, as in engineering, designing a creative problem-solving process that is open, sensitive to process and timing, flexible, and responsive. Nonformal reason provides the resources that can rationalize processes for investigating, discussing, and resolving moral problems in this way. Whitbeck’s characterization of moral problem solving as akin to engineering design provides the right way to conceptualize moral problems and to understand how nonformal reason provides the resources to deal with them. And whereas engineering problems are replete with potential and actual compromises, and practical ethics problems likewise, so that it is an essential part of rational method for problem solving, compromise has no place in formal moral philosophy. 6.2  The Deduction-from-Principles Account of Compromise: A Critique There are two logically respectable ways of resolving moral conflicts. One is to revise the applicable obligations that are inconsistent to make them consistent and then deduce what ought to be done from the resulting set of consistent obligations. The other is to identify a consistent set of obligations that is deemed to be more compelling in the particular circumstances and then deduce what ought to be done from the preferred set of more weighty obligations. These two strategies, described respectively as specification and balancing, are adopted in the current practice of moral philosophy for resolving conflicts among principles. Insofar as the result of either method can be called a compromise (a matter discussed later), it is a conflict-removing compromise. Conflicts must be expunged because they are logically intolerable. We argue that this orthodox philosophical approach to moral conflicts, and whatever conception of compromise it might yield,

Ethics as Design and Its Two Distinctive Methods 

147

are not an accurate representation of either the nature of, or the resolution of, most ethical dilemmas, nor do they support our moral conceptions of a morally mature person. Specification3 renders applicable principles consistent by modifying the formulations of the principles to reduce their ambit or scope (in logical terms, the extensions of the principles). The statement of one or more of the conflicting principles is amended to restrict its applicability so that the principles no longer conflict—for instance, in the way the conflict between relieving suffering and preserving life in Ms. B’s case (see section 5.2.1) was removed by restricting the latter to nonterminal conditions. Because the ensuing restricted obligations that apply now apply fully, there is no conflict, and what ought to be done is clear. But how is this conflict-removing restriction to be understood and justified? That depends on the underlying epistemology. With foundationalism, if the restrictions imposed can be justified by deducing them from more basic ethical principles, conflicts can be resolved without the need for any extralogical reasoning. If not, the only resort is to an a posteriori discovery of an a priori truth—to claim that the newly restricted statement of the principle simply corrects a hitherto mistaken formulation of the principle and reveals what the true content of the principle always has been. With coherentism, the only option is to invoke a notion of coherence or fit, with reflective equilibrium being the preferred philosophical candidate. We consider these options below, but first turn to the other familiar strategy, balancing or weighing. Balancing or weighing is the approach of moral intuitionism. Applicable principles that conflict vie to be the sole principle, or priority principle, to be applied. Principles are regarded as being prima facie reasons for a decision, and intuition is used to “balance” or “weigh” inconsistent prima facie principles to determine which should prevail in the circumstances. Balancing or weighing eliminates logical inconsistency by rejecting the application of the prima facie principles deemed weaker in their particular contests with other morally relevant principles. Both the strategy of specification and the strategy of balancing or weighing encounter three problems. The first problem is how to provide a clear sense of rational justification for the favored resolutions of conflicts. Consider first a foundationalist approach. That there is a problem is obvious for balancing and weighing because those techniques appeal solely to intuition, but intuitions, however they are understood, are notoriously variable across agents and modifiable across experiences. Given that diversity, how are conflicts between inconsistent intuitions to be rationally resolved? The intuitionist has no response. The specificationist who appeals to a

148 

Chapter 6

posteriori discovery of a priori truth is on no better ground: the mere insistence that the newly restricted formulation of a principle simply represents the principle as it always has been is no more rationally compelling than an intuition. The claim that restrictions can be deduced from more basic principles at least offers the possibility of a rational justification. Its problem, however, is that providing those deductions has proven to be very difficult, if not impossible. Deductive rationality is stuck with consistency and inconsistency as its only relations among principles, so on what grounds could one moral principle be shown to give in to, or make concessions to, another moral principle? The only recourse would seem to be an appeal to second-order moral principles that prescribe the applicability of firstorder moral principles, but such principles are neither readily forthcoming nor have much moral force. Indeed, any such second-order principles seem post hoc, presupposing and endorsing the hard work of moral consideration rather than enlightening it. A coherentist approach has the opposite problem. Coherence is readily attainable when it means no more than logical consistency, but logical consistency is too weak to provide any meaningful justification of restricted principles—any two completely unrelated propositions are logically consistent. So consistency could be attained simply by adjoining any consistent restrictions to the conflicting principles. Recognizing that, Henry Richardson (1990) proposes that it is possible to justify restricted principles by incorporating them into a coherent moral theory so that they fit together with other such principles. But again, the notions of “coherence” and “fit” remain refractorily vague and nebulous. Intelligible coherentist consistency is easily attainable, but its weakness provides no meaningful rational justification.4 In sum, within traditional moral philosophy there are no readily evident ways out of this first problem. The second, and related, problem is that without a rational justification for the restriction or intuitive dominance of a principle, the parties to a compromise are left with their moral integrity violated. The third problem is the most revealing: both specification and intuitive dominance fail to capture the ongoing claims of unsatisfied moral obligations, which both characterize our moral condition and drive our moral development. For instance, they discount the struggle of the single parent, where both obligations (to a child and to work) continue to bite and foster regret,5 and similarly fail to do justice to the many other intractable struggles at the core of our moral lives.6 This third problem also exposes the fundamental conundrum for the notion of conflict-removing compromise: How can a conflict-removing

Ethics as Design and Its Two Distinctive Methods 

149

compromise be a compromise?7 If there is no conflict, there is nothing to compromise. So where might any putative compromise be located? Specification produces a consistent set of restricted principles as outcome, so no compromise can lie there. Perhaps the compromise lies in the restrictions that curtail the obligations. But it cannot be there either if the restrictions are to be justified as really always having been there, and especially if they are justified by deduction from more basic ethical principles, since the restrictions then would be a fundamental part of the principle and not a compromise of it. Restrictions justified in this way would render compromise no more than an initial illusion that the principles were unrestricted. The same holds for intuitive balancing and weighing. Compromise cannot lie in the outcome because that is a single dominant principle, and it cannot lie in the restrictions if these are to have moral validity, since they have to be eternal moral truths discovered by intuition. Compromise again becomes no more than an initial illusion that the principles were unrestricted. In sum, while conflict-removing consistency provides a basis for moral action, vindicating the validity or rationality of the outcomes of specification or intuitionism eludes standard moral philosophy. Establishing the rationality of compromise requires a richer understanding than this of what compromise is and how compromises are formed and where they lead. It requires the realization that recognizing, and a fortiori creating, an intelligent genuine compromise is a matter of skilled judgment formed and guided by nonformal reason. This realization—and radical departure from orthodox moral philosophy—sets the stage for reconceiving the nature of compromise and its role in moral life, along with reconceiving the conception of rationality that underlies compromise. 6.3  Moral Compromise Compromise is a pervasive fact of life.8 It is occasioned when the several obligations we acknowledge in a situation cannot all be fully met and we cannot avoid the conflict because inaction will generate a worse outcome. Compromise occurs, for example, when a single parent cannot both satisfactorily parent a young child and earn the income that living together demands; when a doctor cannot both preserve a patient’s life and control pain or respect a wish to die. War zones are permanent sites of wrenching moral compromises, but they differ from political or commercial arenas only in degree, not in kind. For people striving to lead moral lives, compromise is always unsettling, and it can be devastating. Yet compromise can also be uplifting. Think of the single mother who determinedly and

150 

Chapter 6

skillfully picks her way between caring and working, committed to both and faithfully bearing the extra effort of feeling and thoughtfulness. She is an exemplar of moral and emotional maturity, someone who has grown, not shrunk, through the experience of compromising (without in any way reducing her preference to eliminate the compromise). Would-be conflicts that are removable do not involve compromises. Suppose a doctor, faced with an aged patient in extreme pain, concludes that the patient can be made sufficiently comfortable to satisfy the obligation to control pain without the indicated high doses of a drug that would accelerate the demise of the patient. Then there is no need to compromise; each obligation can be satisfied. Or suppose the doctor concludes that the obligation to preserve life does not apply to terminally ill patients for whom death is imminent but only to patients who have the prospect of continued life. Here, too, there is no need to compromise; one obligation does not apply, and the other is fully met. But this is not the situation of the single parent. The obligation to bond with and nurture the infant is ongoing and pressing but is practically defeated if there is no bed to sleep in, no food on the table, no time to play. Likewise, the obligation to earn a living is ongoing and pressing but is defeated if she is never with her child. Neither can be set aside, and neither can be fully met. Moreover, when each is met to the extent it can be, the demands of the fully applicable but partially unmet obligations continue and will be keenly felt. The parent remains acutely aware of the time not spent with the infant and the needs and advantages of life not provided, as well as of the career opportunities foregone. This is the stuff of real, conflict-preserving compromise because the forces of the competing obligations remain undiminished or unrestricted throughout, yet each is partially defeated in the outcome, to a degree deemed proportionate to its proper role in the current circumstances. It might be tempting to argue that when individuals inherently cannot satisfy all the obligations they recognize, then, since “cannot” implies “not ought” (since “ought” implies “can”), they are not under an obligation to do what they inherently cannot do. That is so, but it does not release them from compromise. Although they are not obligated to act so as to meet all their acknowledged obligations, it is quite consistent with that constraint that they are obligated to continue to acknowledge (to “feel”) the press of all the obligations on them. That is in fact what happens, and it is crucial to our being a moral person that it should do so, for it keeps us engaged with the entire moral order. And that engagement in turn provides the widest basis for appreciating new options for improving the situation as they arise. (See also section 6.3.2 below.) If we are not to abandon these

Ethics as Design and Its Two Distinctive Methods 

151

unmet obligations to some limbo of unacknowledged moral experience, it is important to acknowledge a legitimate place for compromise in our lives. In doing so we should not confuse two distinct kinds of “compromise.” There are compromises of principle that are immoral, for instance, accepting a bribe to allow an illegal act. If a proffered resolution of a conflict among obligations were to include such a compromise, moral integrity would be violated. Those are corrupting compromises. But that problem is quite distinct from the issue of whether conflict-preserving compromises per se are immoral or jeopardize moral integrity. We must hope not, since they stand at the core of our moral experience and development and are ubiquitous, found not only in every private life but also in every public policy, every design process, and so on. The goal, then, is to find a proper role for noncorrupting compromise in the moral life and a proper rationale for that role. 6.3.1  The Nature of Compromise Martin Benjamin provides an instructive example of compromise at work in a moral conflict in an intensive care unit (ICU) (Benjamin 1990, chapter 2). A young, single woman who has suffered severe brain damage from viral encephalitis has been admitted to the ICU, and her overwhelmed parents have ceded decision making to the medical staff. An experienced critical care nurse and the attending physician in the unit disagree about whether aggressive treatment should be continued. Other members of the staff also disagree about how the young woman should be treated. The head nurse argues that it is highly unlikely that she would want to be kept alive in these circumstances, that even if with treatment she would live, she would be likely to have multiple, severe, debilitating deficits, and that continuing aggressive treatment is very expensive and not worth the cost given the exceedingly low probability of obtaining significant benefits when other patients could benefit significantly from those resources. The physician believes that the young age of the woman, the sudden onset of the disease, and the patient’s previously excellent health suggest that if anyone could recover, she could, and that recovery is of most benefit to the young. The physician also appeals to the inherent value of human life and the importance of remaining steadfastly dedicated to preserving and prolonging human life. Because the nurse and the attending physician are at an impasse and the staff is divided, the status quo—aggressive care—continues. For many reasons, including putting an end to a distressing, demoralizing stalemate, a resolution is needed, perhaps a compromise. Benjamin proposes an account

152 

Chapter 6

of compromise that has three components: an explanation of the circumstances in which searching for a compromise is viable; a description of the process of compromise; and an analysis of what is required for the outcome of a process of compromise to be a genuine, or strict, compromise. The circumstances of compromise consist of factual uncertainty; moral complexity; a continuing, cooperative relationship between the parties; an impending, nondeferrable decision affecting both parties; and scarcity of resources. The more of those features that are present, the stronger the case for a compromise. Whether to pursue a compromise is a matter for nonformal judgment. A process of compromise includes an exploratory discussion among the differing parties, marked by mutual respect, a fair hearing of argument, and a willingness on both sides to respond to “new information, insights, arguments, or understanding” (Benjamin 1990: 32–33). A genuine compromise results when substantive differences between the parties continue, so that for there to be a resolution the parties have to agree to split their differences, each side giving up something. For instance, the parties could agree to continue aggressive treatment but only for a specified time, after which they would assess the patient’s condition using criteria they have mutually accepted. If by that tine the patient’s condition was improved, aggressive treatment would continue. If it remained the same or deteriorated, less aggressive treatment would be instituted and more importance would be given to the equitable and efficient use of the resources for treating other patients (34–35). This is a genuine compromise: although agreeing to the outcome, does the nurse still regret the other patients’ losses of support, and does the doctor lament the violation of the obligation to vigorously pursue life? Surely yes and yes. Compromise is the reluctant bending to the strictures of finitude in the face of the continuing matrix of acknowledged obligations. There is no action that can be taken by finite agents in this world, individually or in a group, that can satisfy all our values, and so each act is in principle a conflict-preserving compromise, even if a minor one in the wider compass of life. 6.3.2  The Moral Life of Finite, Developing Agents Finitude means that we begin in ignorance—ignorance not only of the substance of a good life, but of the terms in which to conceptualize it and the methods to use to validly acquire it—and that we possess limited resources, including reasoning capacity, with which to pursue relief from these limits. For finite agents, life is full of conflicts among valuables, whether those valuables are goods all of which cannot be obtained or obligations all of which cannot be fully satisfied. Compromise derives from our finitude in

Ethics as Design and Its Two Distinctive Methods 

153

the face of the complex, multifaceted character of valuing. To return to our earlier example, single parents properly value both caring for their children and working, but their finitude prevents them from fully satisfying the obligations of both. This conflict is contingent insofar as a parent’s need to work could be removed by a benefactor (governmental or personal) or a lucky accident (winning a lottery), which makes clear how finite monetary resources underwrite it. But this feature alone misses the intrinsic value to a parent-as-social-participant of working, developing work-related skills, taking pride in occupational accomplishments and so on, and the equivalent social dimensions to parenting (developing play-group skills, and so on). In this regard the finitude constraint is more fundamental: finite creatures must learn by social interaction at a specific time and place, but caring and working, then, are at least attention-competing, and typically spatiotemporally mutually exclusive, activities. Similarly, a nurse’s capacity to care remains finitely distributed among patients even with expanded resources, and to propose increased staffing as a resolution only promotes conflicts about the allocation of finite resources to a broader social context.9 Coping with moral ignorance and conflict is the essence of the moral life of finite agents, and Betty Winslow and Gerald Winslow astutely identify the implications and the dangers of this plight: To the extent that two or more of our own firmly held moral values are in conflict, we know for certain that any decision we make will represent a loss of something important to us. We also know that we could be wrong. The complexities and the uncertainties of the matter render us less than completely confident about our choices. This acknowledgment is nothing more (or less) than the appropriate level of humility, given human limitations. Such humility helps us to remember the value that was lost when its competitor was given priority. It is often tempting, of course, to make the moral life simpler than it really is by pretending that one of the values in conflict is not so valuable after all. But this approach to eliminating moral perplexity is neither honest nor prudent. (Winslow and Winslow 1991: 318)

Here again the abstract simplifications of moral philosophy are not faithful to our moral lives. Rather than representing a minor and unfortunate aspect of life to be eradicated, as the logical consistency approach would have it, conflicts among values stand at the core of our moral lives, since there is nothing of moral substance a finite agent can substantially improve without resorting to compromise.10 At least as importantly, living with conflict-preserving moral compromise is what drives our moral development. As we noted earlier, the single mother who effortfully navigates between caring and working, and

154 

Chapter 6

willingly bears that effort, exemplifies moral and emotional maturity, having grown in the process of compromising. Moreover, the continued conflict of values means that she remains alert for opportunities to reduce the remaining conflict, for example, by acquiring a job with on-site child care facilities or child-friendly hours, and/or co-opting a family member to assist in caring, and/or campaigning for such employment and other improved sole-parent support. Both as individuals and as a species we have to work our way laboriously toward improved moral knowledge and practices; that is one of our greatest developmental tasks.11 Moral development proceeds substantially through the experience of conflict resolution—it is our key gateway to moral improvement.12 We have already noted that the ongoing matrix of acknowledged obligations leaves us permanently open to revisiting and improving our past decisions in the light of our accumulating experience with relevant succeeding decisions. But there is another aspect also worth noting: the construction of a richer coherence. The continuing pressure of the full range of acknowledged obligations ensures that creative construction will search for the richest ways possible to do justice to all obligations, rather than just those obligations met by the momentary compromises we are forced to make. This role of creative construction is crucial in overcoming the fragmented normative myopia that might otherwise ensue from the trail of constrained practical actions we must take. In chapter 9 we provide an extended case study of a prominent health care policy issue, the allocation of scarce kidneys from deceased donors for transplantation, where just this process plays out in a surprising way that can only be understood as a compromise. It is the history of improved resolutions to problems of compromise, typically through ways that lessen their constraints, that gives force to our guiding idea of ethics as design for flourishing. Finally, the continuing pressure of the full range of acknowledged obligations provides the proper basis for extending and enriching the compromise policies we have. The full panoply of norms in force orients us normatively to new situations; we enter a situation alert for how our compromised but still present values might play out in the new circumstances. It is the continued appreciation of the undiminished obligations that guides the adaptation of solutions to the particularities of other contexts. For example, because of the exploration that the women who received genetic counseling did with respect to their families’ core values, adaptive capacities, and, yes, limitations on these as well, they were equipped to respond to a wide range of challenges to the futures of their families, from caring for variously handicapped offspring to moving to a new country. As situations change,

Ethics as Design and Its Two Distinctive Methods 

155

the claims of some values will be more easily or comprehensively obtainable, while the claims of others will recede—for example, because of higher associated costs—so compromises among them will be re-struck. Doing that renders ethical decisions context dependent. This conception of a normatively oriented agent, forced to continually compromise while accumulatively learning through that ongoing process what is possible and rationally preferable across the terrain of practical situations, provides exactly what is needed to understand the developmental process of becoming both a mature person and a wise actor.13 From this developmental perspective conflict-preserving compromise is in itself no assault on integrity, as Benjamin feared it would be. To require that conflicts be removed is to set the bar for moral performance above what is in principle achievable for finite agents, which is incoherent. And, we have insisted, it would be to miss the real moral quality to be had through struggling with compromises. Indeed, a person would be considered to lack integrity if, striving to reduce the stress of unsatisfied acknowledged obligations, they erected the fiction that for each conflicting pair of obligations one did not apply. Moral integrity is not the mere preservation of consistency. Moral integrity is an ideal to be pursued, the virtue that, in their discussion of compromise and integrity in nursing ethics, Winslow and Winslow describe as “constant sincerity of moral purpose” (1991: 309). As finite agents we are perpetually caught in moral conflicts, and the best we can ever do is satisfy obligations within our available resources and to the degree they are important in the specific situations in which they arise. For a finite agent, doing that precisely is having integrity. Virtue is learning to judge wisely which obligations are relevant, what their comparative importance is in a specific context, and which kinds of compromise are best overall. We do not even see any a priori case for inherent patterns of priority among conflicting general obligations, as the deductionfrom-principles approach might like. Life may just present us with a set of situations wherein every combination of priorities appears somewhere in best solutions—and yet we can have moral integrity. Thus we conclude that conflict, and how it is handled, especially in compromise, should be embraced as being central to identity and moral integrity. In this regard it is instructive to review Benjamin’s struggles over moral integrity. Benjamin holds that a conflict-preserving compromise threatens moral integrity because the parties to the compromise agree to act in a way that does not fully uphold their obligations. If both the nurse and the doctor regard the outcome of the ICU process as superior to their initial positions and adopt it for that reason, there is no compromise and no worry

156 

Chapter 6

about their integrity having been violated. But if they continue to differ, for example, about the relative weights to be attached to saving life at all cost, the probability of success, and the fair and effective allocation of scarce resources, and remain convinced of the correctness of their initial positions and agree to the outcome only as a practical, expedient way of ending a disagreement that has to be ended, then there is a worry about compromising their integrity, not just their particular moral principles. As Benjamin puts it, “moral wholeness or integrity, it would seem, is incompatible with their agreeing to … a compromise outcome” (1990: 36). His response to this worry is, in the ICU case, to broaden the terms in which the problem is framed to include the participants’ professional obligations. The enveloping issue then is how the parties to the disagreement can manage their disagreement in a mutually respectful manner that will allow them to maintain amicable relationships and continue to work together productively in the ICU and, at the same time, emerge with their personal integrity intact. Benjamin quotes a passage from Arthur Kuflik to make this point: [T]here is more to be considered than the issue itself—for example, the importance of peace, the presumption against settling matters by force, the intrinsic good of participating in a process in which each side must hear the other side out and try to see matters from the other’s point of view, the extent to which the matter does admit reasonable differences of opinion, the significance of a settlement in which each party feels assured of the other’s respect for its own seriousness and sincerity in the matter. (Kuflik 1979: 51, quoted in Benjamin 1990: 36)

In this regard the problem for the nurse and the doctor in the ICU is like the problem for K’aila’s parents (section 5.2.3): they are deciding not just for someone who cannot decide for himself or herself but for themselves as well. The problem must be framed in a way that includes that crucial dimension. Benjamin believes that framing the problem more broadly defuses the worry about integrity. The kinds of considerations that Kuflik identifies, he points out, reflect values and principles that many of us hold dear and that partially determine who we are and what we stand for. If, therefore, we suppose that … [the nurse and the doctor] also place a high value on tolerance and mutual respect, it is not so clear that agreeing to the proposed compromise constitutes a threat to their integrity. On the contrary, taking into consideration all of their values and principles plus the fact that they disagree and that they are in circumstances of compromise, the compromise solution may be for them more integrity preserving than any available alternative. (Benjamin 1990: 36–37)

Ethics as Design and Its Two Distinctive Methods 

157

If the nurse or doctor is troubled about compromising her principles in deciding what should be done for the patient, she can assuage that concern by realizing that she is not compromising her principles about what kind of person she should be and how she should handle disagreements with others. She remains open-minded, fair, respectful, and tolerant. This reformulation of the problem does not, however, remove the threat to integrity that compromise poses. Benjamin is right to introduce considerations about what kind of person the nurse and doctor ought to be because doing so commendably recognizes that moral problems are people’s problems, and he is right that this compromise does not violate the integrity of the nurse or doctor, but he is wrong about why the compromise does not violate their integrity. As morally relevant as the new considerations are, they do not, as Benjamin supposes, eliminate the threat to integrity. Rather, they shift the locus of the compromise. The two parties now have to make a comparative judgment about how important their obligations to the patient are and how important their obligations to themselves are. If the initial conflict-preserving compromise about the good of the patient jeopardized their moral integrity, so does an expanded conflict-preserving compromise about the good of the patient and their own good. Benjamin’s myopia likely has the same source as his fear about preserving integrity, namely, his tacit assumption of the deduction-from-principles account of morality, where the logical inconsistency of principles brings moral incoherence. Recognizing that morality is much more expansive and much more complex than reasoning from principles is what removes the threat to integrity. Indeed, one might worry more about the integrity of those who too easily claim to reach consistent positions for too easily avoiding conflicts and the labor of conducting themselves with unceasing fidelity to all the dimensions of their acknowledged obligations. The compromise between the nurse and the doctor encompasses the full panoply of relevant moral considerations and the continuing salience of the unfulfilled obligations. The two parties will regret what they ought to do but cannot do for this patient and their other patients, and they will regret the respects in which they ought to be but cannot be better persons. At the same time, their regret will give them the awareness and the motivation needed to search for ways to avoid or mitigate similar conflicts in the future—to improve what they do and how they do it for the benefit of the patients and the staff in the intensive care unit.14 By continuing to acknowledge the entire matrix of obligations, despite having to compromise in acting, persons retain their full normative orientation to the world and their full awareness of the compromise made, and so they develop

158 

Chapter 6

morally and are continually primed to do better in the future as circumstances, including self-made circumstances, permit. 6.3.3  The Rationality of Compromise Resolutions of moral problems will not preserve moral integrity if they are irrational. But how can it ever be rational to compromise basic values? When there is no morally preferable and practicable alternative, as in the ICU case and as happens often in our lives, how then do we go about rationally constructing and accepting a compromise? The model and practice of nonformal reason presented in chapters 4 and 5 provide the resources, and the ICU case is an instructive example. Utilizing nonformal reason in a process of compromise involves the joint employment of its four resources—observation, formal and informal reasoning methods, creative construction, and comprehensive critical assessment. Benjamin’s account of the ICU process, though brief, clearly implies the use of these tools, from the clinical observation of patients and the pragmatic observation of resource scarcity, through the creative construction of anticipative scenarios for the competing uses of the scarce resources, the reasoning about the likely satisfaction of values in those scenarios, and the critical analyses of the ways the scenarios compete; from which among possible compromises are leveraged, to the critical, reflective review of the overall situation to ensure clarity, comprehensiveness, and lack of bias. How all these resources are brought to bear in a detailed and mutually integrated way cannot be reduced to any simple procedural rules or specified in advance; their effective and productive use is a matter of the wise design of a process for the circumstances, and that too requires judgment.15 In particular, there are many places in the ICU deliberation process for pursuing reflective judgment. It can be used to establish the scope and nature of the problem, for example, whether the disagreement requires a compromise and, if so, where it lies on a spectrum of relations from conflicted to cooperative among the participants and what values are implicated, not only in the focal situation—providing treatment to the ICU patients—but in the relationships among the participants in the decisionmaking, the members of the ICU staff. A mature appreciation of a problem can emerge only after substantial discussion of all the various aspects of the situation. In the ICU case that examination includes the clinical prognoses of the ICU patients and the best treatments for them, the available resources and whether they might be feasibly augmented and, if so, at what cost, the likely consequences of compromising the indicated treatment for each patient, and so on. Then questions of which treatment scenarios are

Ethics as Design and Its Two Distinctive Methods 

159

most salient, what values are at stake in each scenario, and how these values might be “traded off” have to be addressed. Each of these matters is subtly complex, and the three are interrelated. The aptness and thoroughness of their consideration contributes centrally to the rationality of the overall process. A process that is wide enough to encompass the range of relationships involved in practical moral decision making, and which incorporates MEU as simplified sub-cases as appropriate, is wide reflective equilibria (WRE). The notion of narrow reflective equilibrium adheres to logical processes alone to achieve integration, so it needs to be transformed into a reflective process that utilizes the resources of nonformal reason to create rich reflective practices capable of supporting an innovative problem-solving process. This is the subject of the next section. 6.4  Reflective Equilibrium Liberated, Extended, and Enriched John Rawls’s initial formulation of a decision procedure for ethics (1951) and his subsequent conceptions of a narrow reflective equilibrium and a wide reflective equilibrium (1971, 1974) provide an introduction to, and motivation for, liberating reflective equilibrium from not just formal reason’s foundationalism but formal reason’s demand for justifications of right answers,16 by extending the width of reflective equilibrium and enriching reflective equilibrium to encompass moral compromise. Extracted from the prevailing philosophical and epistemological background, reflective equilibrium can focus on the learning process rather than the features of its outcome product, in particular it can employ the resources of nonformal reason. The resulting new regulatory conception of reflective equilibrium has the resources to be far more powerful and far more helpful. And it can accommodate, as a formally degenerate version, formal logic-based philosophy’s account of reflective equilibrium. In this section we examine wide reflective equilibrium (WRE), now embedded in nonformal reason, as a key tool in rational ethical decision making. 6.4.1  Reflective Equilibrium At its root equilibrium is a stasis among entities that pull in different directions, whether those entities are literal physical forces or metaphorical social pressures or normative moral obligations, with the additional condition that any movement away from the stasis creates a restoring force that returns the system to it. A simple physical case is that of a marble at the bottom of a round-bottomed bowl: the marble is at rest in the center of the

160 

Chapter 6

bowl’s bottom, and any movement away from there involves climbing up the side of the bowl whereupon gravity returns the marble to the bottom. In this case the marble at equilibrium is motionless and the equilibrium is static, but there can also be dynamic equilibria in processes where a stasis in some features of the process can be restored if disturbed and all the while the process continues, indeed its continuing is typically what provides the restorative action. Standing waves in a river rapids provide a simple physical example. Almost all the physical equilibria of functional interest in living systems are dynamic equilibria that leave the underlying system processes (metabolism, memory updating, and so on) running. Equilibria in social-pressure metaphors are dynamic. With political equilibria, for example, all the underlying social processes, from bread baking to zoo keeping, are typically presumed to be running while, for instance, opposing parties may reach an equilibrium formulation of an education policy such that all options to depart from that stasis generate net incentives to return to it. Similarly for equilibrium among moral obligations. All the other aspects of personal and social life are presumed to be proceeding in a context while an agent locates an option from which there is no net reason to depart and thus reaches an equilibrium in their commitments. In arriving there the agent has carefully considered each of the countervailing obligatory considerations—that is, the values whose pursuit would, taken separately, move the agent away from that option—with respect to all the relevant consequences and manners of trade-offs among them available in the context. Hence the term reflective equilibrium. Reason demands reflective equilibria for all human commitment equilibria, individual or social, since anything less implies a relevant factor that has not been taken into account, and this violates at least the systematic critical assessment requirement of nonformal rationality. Of course, equilibria hold only as long as their inputs hold. Changes to a moral situation—for instance, new information that alters ethically relevant characteristics, or new values that are injected from legislation or from new institutional responsibilities—will reopen the matter, with the present equilibrium often leading to a new reflective equilibrium. 6.4.2  Wide Reflective Equilibrium The decisions of Ms. F about her death (chapter 2), of the ICU staff about continuing aggressive treatment for a young encephalitis patient (section 6.3.1), of the women at risk of having a child with a genetic condition (chapter 1), and of the Kidney Transplantation Committee about the allocation of kidneys from deceased donors for transplantation (chapter 9) affect different

Ethics as Design and Its Two Distinctive Methods 

161

numbers and ranges of people, but that variability is not itself ethically significant and is not what we mean by the width of equilibrium. Rawls’s initial method of reflective equilibrium worked back and forth between considered moral judgments and principles to achieve coherence between the two. That narrow reflective equilibrium evolved into a wide reflective equilibrium for justifying a moral theory by reaching coherence among a much wider array of considerations in the same back-and-forth manner between considered commitments. Norman Daniels explains that wide reflective equilibrium is superior to narrow reflective equilibrium because it marshals “the broadest evidence and critical scrutiny,” encompassing beliefs about particular cases; about rules, principles, and virtues and how to apply or act on them; about the right-making properties of actions, policies, and institutions; about the conflict between consequentialist and deontological views; about partiality and impartiality and the moral point of view; about motivation, moral development, strains of moral commitment, and the limits of ethics; about the nature of persons; about the role or function of ethics in our lives; about the implications of game theory, decision theory, and accounts of rationality for morality; about human psychology, sociology, and political and economic behavior; about the ways we should reply to moral skepticism and moral disagreement; and about moral justification itself. (Daniels 1996: 6)

The core idea of WRE is that the greater the range and diversity of the considerations that underlie a particular reflective equilibrium, and the more those considerations mutually complement and support one another, the more reasonable it is to rely on that reflective equilibrium as a rational basis for action. The kinds of beliefs Daniels adduces pertain more to moral theory and moral philosophy, however. Developing WRE for a particular moral problem is a somewhat different undertaking. In that regard, return to the ICU staff’s disagreement about continuing aggressive treatment for the young woman with encephalitis. Complementing the ICU staff with the staff of an IDU (infectious diseases unit) would be of little ethical significance; it would add numbers but would not expand the relevant medical expertise or ethical experience. We already have seen, though, how diversity between doctor and nurse in this case can significantly test the ethical options, as we have seen the diversity at work in addressing the plights of the children dying from leukemia (section 5.1). The different roles of doctor and nurse provide different insights into patient care and generate different priorities among values, perhaps even some different values. Together the wise judgments proceeding from both roles provide a more nuanced and critically tested exploration of options.

162 

Chapter 6

Another relevant diversity worth pursuing would be to search the ICU for occasions when trade-offs have been made between the values of preserving a young life and maximizing quality-adjusted life-years (QALYs) across other patients, to see what considerations were entertained, what arguments were offered, and how and why decisions were made, and whether there were any pertinent specific differences in those situations.17 Other trade-offs between one and many, young and old, and high risk– high reward and low risk–low reward might also be consulted. And perhaps especially relevant, past compromises reached by the ICU staff (other than the temporarily-support-then-reassess sort) could be sought, and, if found, assessed for their appropriateness and their ethical and medical suitability in the current case. If agreement on a course of action in the ICU emerges from such a network of investigations, the connections and similarities discovered will mitigate the distress of the staff and perhaps strengthen an evolving WRE about a clinical and moral approach to troubling ICU decisions. Alternatively, the outcome might convince the ICU staff that in some circumstances they should not adopt practices from elsewhere. Nevertheless, with a risky or innovative decision, wider coherences could possess increased value because they provide the most similar guiding decisions. In fact, for each moral issue in a context we can associate a landscape of potentially relevant decisions from which to learn—from the ICU on other occasions, from other ICUs, from other institutional components like IDUs, from attached ethics counselors, from other hospitals, provinces, and countries, and so on. A sound equilibrium will include an account of which of these decision sites was visited and informed the current WRE process and which were not, for whatever reasons (time, resources, relevance, and so on). Whatever decision is made, however, a wide search for relevant, helpful information and direction will have provided evidence and argument that will have clarified, and ultimately supported the decision. The support will derive from gathering positive analogies with other ethical problem situations, modified by relevant negative analogies from these sources, all judiciously refashioned as learning continues. For example, a current situation may have as its ethical focus allocation of a scarce resource to a queue of people. Then the primary positive and negative analogies will be found in other such queue-servicing situations. For instance, the values that dominate in rules for serving queues for concert tickets and airplane boarding are different from the values that prevail for emergency exiting from an airplane and for surgery (life saving or not). Comparing these cases sensitizes assessment to which features ought to drive the ethics of a situation.

Ethics as Design and Its Two Distinctive Methods 

163

For instance, the case for continuing aggressive treatment for the young woman in the ICU might gather strength from a variety of arguments for equal opportunity to experience life (similar to priority for concert tickets) but be weak in the context of compelling arguments favoring equality of access to treatment (similar to having disabled persons board an airplane first). Any shifting of analogies, for example, from concert tickets to airline queuing, whether considered positive or negative, requires moving the ethical focus of the situation, and that must also require at least disambiguating and usually reformulating how the ethical situation is conceived, and hence what constitutes a valuable solution, and so on for all of the features involved in the cognitive problem-solving process (section 4.3). Webs of analogies produce much stronger mutual support relationships between problems-with-their-solutions than mere logical consistency between them because the webs identify a common substantial basis for approaching web-related problems-solutions, one that has survived critical scrutiny and generated satisfactory solutions elsewhere and that has ultimately led to a valuable solution in the current case. These bonds form fundamental parts of the formation of a WRE. Indeed, the bonds reach out across ethical problem situations to form part of a pre-WRE web of mutually bonding ethical analyses that act as extended guides to WRE formation and ethical decision making, just as they do within case law and (less formally) in crime detecting procedures. At various times a perceived unity to science, physics especially, has been the inspiration for constructing moral theory from which ethical action can be deduced; and this, ironically, despite the usually accompanying claim that ethics is very different from science. But a nice axiomatic organization to theory, assumed of science, turns out to characterize neither science nor ethics. It does not characterize science because theories are essentially idealizations—simplified versions of reality—and such idealizations can conflict, even when they are derived from the same underlying principles, because they use different idealizing principles.18 For example, models of hydrodynamics near boundaries and in midstream free flow are incompatible, yet the two approximately converge in intermediate conditions. Molecular and physiological cellular theories are similarly related. Even the current foundational theories of physics—relativity and quantum theories—present incompatible characterizations of dynamics, the one for the very large (relativity) and the other for the very small (quantum), and are plausibly incompatible for similar reasons. The simple axiomatic conception of unity certainly does not characterize ethics either, with its extremely diverse and often mutually conflicting fundamental principles and values.

164 

Chapter 6

But the notion of webs of mutual support where common principles are used across situations in various subdomains, matching up at subdomain “boundaries” where the same principles analogously apply, does characterize both science and ethics and also law. So in this more complex, multidimensional, approximating sense, both may express a similar sense of unity. There is another dimension to the kinds of unity that science and ethics manifest: in each case a variety of concepts combine to generate a richly textured collection of generalizations and applications, expressing a mutually supporting unity among them. There is, for instance, a body of such responsibilities and protocols developing around the respectful treatment of different cultures, around eco-environmental assessment, and the like. This does rule out mere collections of ideas but, important as that may be, otherwise contributes only weakly to understanding unities. In particular, it fails to illuminate the interplay between support and conflict that characterizes the ideas of science and ethics, or to tell us how to work with support and conflict. Our ethical norms are severally unavoidable, and jointly they conduce to a human flourishing that would be incomplete in the absence of any one of them; that accomplishment is ultimately the clearest way that we know they hang together. And similarly for epistemic norms. Though our norms generate endless conflicts of obligations and often require painful compromises, they combine to allow us to be whole, to support family life, public decency, mutual trust and concern, compassion, the rule of law, tolerance of diversity, and so on, all of which permits individuals to develop into mature, fulfilled persons, that is, to flourish within a genuine culture.19 Our aim in this book is to provide a conception and tools of reason that will at least keep that vision alive and help to understand its nature. 6.4.3  Liberation, Extension, Enrichment Having set out the structure and function of WRE, it is time to turn to its newfound power. The abandonment of foundationalism in moral philosophy liberates reflective equilibrium, first narrow and then wide, from the constraints of formal reason. Relations no longer are a matter of simple formal consistency, and support no longer is confined to deduction-from. That liberation allows reasoning to be extended—to employ nonformal reason’s rich resources of observation, creative construction, multiple reasoning methods, and comprehensive critical assessment in supporting processes that develop satisfying WREs. Following from that, WRE can be extended to include an increased range and diversity of considerations that comprise robust, sensitized, analogy-based support for reflective equilibria. This allows for collective contributions from across the pre-WRE web of

Ethics as Design and Its Two Distinctive Methods 

165

interrelated ethical problem solving to the understanding of appropriate resolutions of each particular ethical issue. Extending WRE to include all the resources of nonformal reason and correlatively locating problem-solving processes at the heart of WRE formation is a first major way that WRE is enriched. That enrichment of WRE is completed with the addition of full-blooded compromise (section 6.2 and chapter 9) which expands accessible solutions to encompass all those that include one or more only partially satisfied values. While positive and negative analogy support, coupled to the resources of nonformal reason, goes some distance toward recognizing and dealing with conflict among principles and values, compromise is the key additional feature that enables WRE to take on these conflicts explicitly and with that to negotiate the rampant complexity of human life. Enriched with the resources of nonformal reason, the pre-WRE web of relationships, and the problem-solving process, compromise is liberated to devise combinations of compromised norms that represent rationally adoptable ethical positions, combinations wherein each norm may be only partially satisfied (to some partial degree and/or in some partial aspect) by the resulting action while its full obligatory force continues undiminished. For instance, chapter 9 provides a remarkable dynamic compromise that satisfies competing norms sequentially over time in a prominent public policy domain. Such compromises manifest flexible context-sensitive adjustment to complex moral dilemmas, preserve pluripotential ethical obligations that continue to orient us to our unfolding world and our life within it, and provide a broad range of options that allows boldly imaginative resolutions of problems to emerge. Together these three features—liberation, extension, and enrichment— constitute a mature form of WRE that can play a powerful role in our private and social ethical lives. WREs become the process through which we form the contextualized equilibria that mark our ethical development as individuals and as societies, and at the same time establish the framework for reworking them when situations and lives change. 6.5  A Shift from Principles to Values We now are in a position to address an issue that has been emerging for some time: the primacy of values over principles. The discussions of compromise and reflective equilibrium take values to be central to ethical understanding and deliberation. It is easy to see why: multiple values can define an ethical situation, whether or not they conflict; values can bear

166 

Chapter 6

conditionally and partially on situations and on the satisfaction of other values, in ways that are multidimensionally supportive or competitive; and so on. Values can be compromised by degrees, which makes it possible to form compromises wherein some or all values are compromised to some extent. Compromises of this sort pervade our lives and are essential to understanding the ethical choices we make. Values bring much more flexibility and power to ethical decision making than do principles, which cannot be applied in such graded, multipolar ways. In addition, the shift to values is congenial to virtue ethics: virtues enter as further values to be satisfied, but with the practical rider that the consequences of their satisfaction are spread across succeeding situations. We argue that these many advantages to values together provide the best reason to adopt values as primary to formulating ethics and that the search for stronger (read: more a priori) reasons for their primacy is a chimera of formalist philosophy. Despite these advantages, values are relatively ignored in orthodox moral philosophy.20 Perhaps this is because the nature and functions of values cannot be captured by logical reasoning with propositions, so values cannot have the rationality of formal reason. Values cannot be assimilated to principles, yet all that is needful of ethical principles can be converted into equivalent values. Consider the ethical principle “Thou shalt not kill.” What underlies it is at least the value of life. The basic obligatory or normative force of a value V is that of promoting V to the maximum extent feasible. Thus the obligatory force of the value of life is to promote life to the maximum extent feasible. While this obligation allows degrees of violation as the occasion may warrant, it also captures the force of the principled prohibition; when the latter clearly dominates, say for unprovoked killing, the value formulation carries the same prohibitive force: it is impossible to promote life when killing for no reason. Moreover, unlike principles, the obligatory force of values remains throughout their practical violations, permitting full-blooded compromises among them. And where principles fail to offer guidance, for example, in situations where someone must die no matter what action is taken, instead of a logical stalemate, satisfying the value to the maximum feasible extent would prompt an exploration of what qualities and quantities of life are involved, what wider impacts on others there might be, and so on. Values do turn up as utilities to be inserted into formal game and decision theory and optimized. But other than imposing the obligation to search for improved efficiency, the utility principle says nothing about the shape of compromises or the construction of equilibria. MEU is compatible with an undiscriminating pragmatism in

Ethics as Design and Its Two Distinctive Methods 

167

which everything is always equally tradable, but it is equally compatible with the richer structure of moral deliberation developed here.21 6.6  Conclusion Two themes of this chapter come together. First, there is Whitbeck’s recognition of ethical problem solving as design (section 6.1). It applies to both the general character of the problem situation—how to bring about a problem resolution of sufficient value while respecting the operative constraints—and the internal problem-solving process, including the key idea of an iterative process in which problem formulation, solution values and criteria, and solution proposals all co-develop. These features exactly define how design theorists see their problem-solving process.22 But what is an appropriate design goal of problem solving? Each solution design creates sufficient value to justify its adoption in the circumstances, but a disaggregated collection of value improvements might lose value by ineptly bringing about incoherence. However, the consideration of culture provides a conception of appropriate unity (section 6.4.2): our ethical norms are severally unavoidable, and jointly they conduce to a human flourishing that would be incomplete in the absence of any one of them. We take the development of such an ethically cultured life to be the goal of ethics as design for flourishing. Second, there is the characterization of nonformal rationality. The discussion subsequent to section 6.1 completes our account of the two tools distinctive to nonformal reason and, with the shift to values from principles (section 6.5), that of the character of nonformal reason as a whole. To be nonformally reasonable is to be committed to applying wherever feasible the resources of observation, reasoning, creative construction, and critical assessment to the improvement of flourishing, expressed as the increasing satisfaction of value(s) and/or increasingly adequate values, aided as appropriate by the formation of compromise and mature wide reflective equilibrium.

7  Designing Deliberation

C

A rational judgment results from a rational process of deliberation, and rational processes of deliberation are constructed, that is, they are designed. Designing deliberation requires rational judgment itself, though, because the endeavor is rife with uncertainty, complexity, and incommensurability. Those who are deliberating want to do what is right, but in a sense of “right” much richer than validity in logic or correctness in mathematics. Judgments that emanate from deliberation have to be right for, and right to, those who are deliberating, not right for everyone. Rational use of the resources of nonformal reason is needed not only in exercising deliberation but also in designing processes of deliberation. The best way to understand designing deliberation is to review various processes of deliberation and the use of the resources of nonformal reason in those processes. Examples from earlier chapters and new examples from social science research are instructive. 7.1  Deliberation as Dramatic Rehearsal The women in chapter 1, at risk of having a child with a genetic condition and deciding whether to try to get pregnant, eschewed the principle of maximum expected utility and instead engaged in the kind of deliberation that John Dewey and James Tufts described in 1909: The notion that deliberation upon the various alternatives open to us is simply a cold-blooded setting down of various items to our advantage, and various other items to our disadvantage … and then striking an algebraic balance, implies something that never did and never could happen. Deliberation is a process of active, suppressed, rehearsal; of imaginative dramatic performance of various deeds carrying to their appropriate issues the various tendencies which we feel stirring within us. When we see in imagination this or that change brought about, there is a direct sense of the amount and kind of worth which attaches to it, as real and

D

170 

Chapter 7

as direct, if not as strong, as if the act were really performed and its consequences really brought home to us. (Dewey and Tufts 1909: 322–323)

These women created and revised imaginative scenarios that were personal, contextual, meaningful, and challenging, and they assessed those scenarios with their intellect and their feelings in a dynamic process of dramatic rehearsal of the kind outlined by Dewey and Tufts. Why did the women deliberate in this manner? Recall that what is important to them is the possibility, not the probability, that they could have a child with a genetic condition, and that possibility is ineliminable. They framed their problem in ways that allowed each woman to develop, refine, and probe her own distinctive situation. They imagine and reimagine what it would be like to live with a child who has a particular genetic condition, how life would be for the child and how life would change for the family and herself. Their imaginative scenarios manifest their accumulated knowledge and experience, incorporate their overt values, and engage their emotions, which express their implicit as well as their overt values. Their deliberation engages their entire being, not just their intellect. That their scenarios are highly individualistic in their themes, particular in their detail, and contextual in their development is as it should be because these women are deciding singularly for themselves, not universally for all women. Throughout this process the women are testing their sense of their ability to accept responsibility for bringing a child with a genetic condition into the world and their ability to cope with a child with a genetic condition. Those are the issues crucial to them, and their process of deliberation is designed to confront them. They have to ask themselves many questions. How complete and realistic are my scenarios? What needs to be changed, and what is missing? What is really valuable to me? What can I compromise, and what can I sacrifice? How genuine and trustworthy are my reactions? Do my reactions change when the scenarios change, and if so, why? What kind of person do I want to be, and do I have the resources to be that person? Eventually the women have to figure out how and when they will stop deliberating. How will they know if they have a green light to try to conceive or if they are not going to get a green light? What if they cannot make a decision, and then come to feel they no longer can do anything usefully? The women need not just to deliberate but to deliberate about how to deliberate. Dewey and Tufts allude to several features of deliberation that contribute to its rationality. One comes from the advantages of mental trials. The consequences of a mental trial are, as they put it, “retrievable,” whereas the eventual “overt” trial or action, which also is a trial of the idea that

Designing Deliberation 

171

produced it, “cannot be recalled” (1909: 323–324). Additionally, because many mental trials can be made in a short time, many diverse, immediate reactions—sensings, recognitions, appreciations, feelings, and emotions— can result. And, as Dewey and Tufts note, when there is a large and broad range of information, “there is clearly much greater probability that the capacity of self which is really needed and appropriate will be brought into action, and thus a truly reasonable happiness result” (1909: 324). There is no more apposite description of what those women deliberating about whether to try to get pregnant were striving to assess than “the capacity of self which is really needed and appropriate.” Another feature is the introduction of Aristotle’s person of practical wisdom, the person who, as Dewey and Tufts put it, is a “good judge of what is really good” (1909: 324). The person who has become virtuous by doing virtuous acts exercises good judgment and displays the proper emotions. And the behavior of the good person becomes “the standard of right and wrong”: There are few persons who, when in doubt regarding a difficult matter of conduct, do not think of some other person in whose goodness they believe, and endeavor to direct and clinch their own judgment by imagining how such a one would react in a similar situation—what he would find congenial and what disagreeable. Or else they imagine what that other person would think of them if he knew of their doing such and such an act. (Dewey and Tufts 1909: 324)

Dewey and Tufts do not claim that imagination, on its own, can determine what rationally ought to be done in a particular situation, but they do recognize rational dimensions of imaginative deliberation: “It helps emancipate judgment from selfish partialities, and it facilitates a freer and more flexible play of imagination in construing and appreciating the situation” (1909: 325). For our purposes, those elements promote nonformal reason’s resources of creative construction and comprehensive critical assessment. They are important contributions to rationality, but what ultimately is needed is the more robust account of rationality that full nonformal reason brings to deliberation as dramatic rehearsal. The design of the process of deliberation as dramatic rehearsal is what renders it rational, in particular, the use of the four resources (chapter 4) in the deliberation. Creative construction begins with an initial general construal of a problem and specifies that problem by imagining detailed scenarios and modifying, elaborating, and revising those scenarios in the process of figuring out what is important and assessing responses to the scenarios. Observation provides crucial, salient information for the creative construction and revision of scenarios. Formal reasoning exposes assumptions and traces the implications of possibilities that are being considered.

172 

Chapter 7

Nonformal reasoning methods such as induction and argument by analogy can relate hypothesized situations to past experiences and others’ experiences and gauge the likelihood and desirability of their occurrence. Comprehensive critical assessment, both intellectual and emotional, scrutinizes, evaluates, and then, as judged appropriate, prompts creative construction of new or modified scenarios, only to be analyzed and evaluated in turn; and similarly prompts revision of operative values, reordering of priorities (for example, between caring and career as components of self-esteem) or introduction of new values hitherto ignored (for example, admiration of disabled people). Throughout this process well-chosen social interactions are helpful, for example, engaging with other parents of a child who might have a genetic condition, caregivers of such a child, and relevant professionals. Such experiences enlarge the range of salient observations, augment the creative construction of scenarios, enhance reasoning methods, and expand the scope and strengthen the force of comprehensive critical assessment. Normative guidance also can come from social behavior that is observed and norms that are invoked, along with requests for help from genetic counselors (although this can be frustrated by the counselors’ perception of their role as nondirective information providers). The resulting overall process is not just rational; it also evinces complete autonomy in the sense that it exhibits all four of Bruce Miller’s kinds of autonomy: free action, effective deliberation, authenticity, and moral reflection (see chapter 2). The customary requirement of making an informed, voluntary decision is only part of the complex notion of autonomy; as the women’s deliberations reveal, their difficulties lie in the other three dimensions of autonomy, the ones that ultimately give substance and meaning to decisions. The best these women can do (and given that “ought” implies “can,” the best they ought to do) is to deliberate in the most thoroughly rational way feasible. Nonformal reason provides the capacity and the resources to do that.1 7.2  Deliberation as Narrative Narratives have been introduced into bioethics as a more practically attuned, helpful way of understanding and working through moral problems. Christine Mitchell attributes their popularity to “the aridness of philosophical ethical theory, especially when applied in the analysis of actual human moral choices” (2014: S12). Much of the motivation for the turn to narratives is the desire to restore persons to the center of morality. Perhaps surprisingly, that goal represents yet another parallel between ethics and

Designing Deliberation 

173

science. Paul Feyerabend, a philosopher of science, explains in the preface to his last book, Conquest of Abundance: A Tale of Abstraction versus the Richness of Being, why and how he tries to connect science to people: I avoid “systematic” analyses: the elements hang together beautifully, but the argument itself is from outer space, as it were, unless it is connected with the lives and interests of individuals or special groups. Of course, it is already so connected, otherwise it would not be understood, but the connection is concealed, which means that strictly speaking, a “systematic” analysis is a fraud. So why not avoid the fraud by using stories right away. (Feyerabend 1999: vii)

Moral philosophers, motivated by their desire to emulate what they take to be the rationality of science, likewise are preoccupied with creating and defending systematic moral theories. So the same question can be asked about ethics. Why not avoid the fraud of abstract moral systematization. Why not eschew logical or conceptual constructions and go right to stories that richly embed people in their problem contexts and link them with other people, both actually and hypothetically, thereby rendering ethics a genuinely practical human endeavor? Doing that is the aim of those who have taken what Hilde Lindemann Nelson calls the “personal turn” in ethics (Nelson 1997: viii). The personal turn is a rejection of the powerful assumptions of logic-based analytic philosophy that ethics is universalizable, impartial, and about right conduct between strangers. Instead, personal-centered ethics has a strong particularist, not a universalist, orientation, although universal principles, often contextually hedged, can be invoked. Universalist assumptions are fostered by formal reason, and they delineate the nature and purpose of the rules that comprise morality. Lindemann Nelson readily concedes that if “doing ethics” is about developing and defending formal moral theories, doing narrative ethics is not doing ethics (Nelson 1997: xii). But if “doing ethics” means “reflecting on the moral aspects of particular encounters within a powerful social institution where what is said and done reveals a great deal about who we are and what matters in our lives,” then doing narrative ethics is doing ethics, “in at least a loose sense of that word” (xii). Her philosophical caution is a perfectly understandable concession to the sway of formal reason. But her concession is unwarranted since narratives provide a richer, more robust, more productive understanding of ethical deliberation that, integrated with nonformal reason, provides a way of deliberating about, and reasoning through, moral problems in a manner focused on people, on encounters and relationships among people, and on the institutions within which people live and interact.

174 

Chapter 7

How is narrative ethics done? Because narrative ethics is diverse and heterogeneous, there is no simple answer to that question. We have to pick an example, and the ethnographic research that Cheryl Mattingly conducted with occupational therapists (OTs) nicely illustrates the coalescence of narratives and nonformal reason. Mattingly objects to the orthodox clinical practice of medicine, or biomedicine as she calls it, because it disregards the social and the cultural, and she traces that marginalization to a by-nowfamiliar source: It is that biomedicine’s very notion of rationality disguises the role of human relations in constructing the clinical problem and arriving at decisions about how to treat it. Distancing from the social and moral may have reinforced biomedicine’s status as a purveyor of truth, but how is this denial and disengagement possible? How can life-altering illness, something that so palpably calls the moral and existential into question, be transformed into a matter for objective rational action, for science and technology stripped of its moral, social, and cultural grounding? (Mattingly 1998: 276)

That reductionism clearly is possible, as Mattingly realizes, because it happens routinely in medicine. Her question is about how the clinical practice of medicine can be effective and ethical when it is no more than the application of scientific medicine. The parallel between Mattingly’s criticism of biomedicine as applied scientific reasoning and our criticism of practical ethics as applied ethics is unmistakable. Both applied approaches strive to deliver the truth that only formal reason can establish, and as a result each produces a constricted conception of, and approach to, its respective endeavor. Just as there is much more to practical ethics than what is recognized by applied ethics, so, too, there is much more to clinical medicine than what is recognized by applied scientific reasoning. Mattingly’s illustration of what is morally excluded from clinical medicine is striking: the “essential moral questions that plague the sufferer of serious, disabling illness—‘Why me. Why now. How can I go on with my life?’ And even ‘Who am I now?’—are precisely the ones biomedicine sidesteps” (1998: 276). Those daunting moral questions are about how it is possible, perhaps no longer to flourish as before, but at least to live well in the midst of profound loss and adversity. As Mattingly emphasizes, as much as clinical medicine wants to evade them, these issues are at the core, not just of what medicine can do, but of the healing that medicine proclaims to be its goal. Mattingly’s alternative vision of clinical practice is exemplified by the storytelling of occupational therapists. Her ethnographic research with

Designing Deliberation 

175

them allows Mattingly to examine “the way clinical practice exceeds the bounds of its own [scientific] ideology and the role clinical storytelling plays in doing so” (1998: 278) and reveals the distinct affinities that this broader conception of clinical practice has with nonformal reason. OTs believe that good treatment plans have to be devised to meet the concerns and needs of particular patients, a goal they describe as “individualizing treatment,” and they regard the ability to individualize treatment as “a highly complex clinical skill” (278). Individualizing treatment is the goal of orthodox clinical practice as well, but the difference is the scope of the individualization. For OTs, exercising this clinical skill involves highly contextualized, flexible judgment that is not just medically but also socially informed and regulated, because individualizing treatment is a matter of judging what is best for a particular patient in a particular situation. The concern to integrate therapy within the real life of patients and their families creates a need to modify treatment in accord with singular life contexts and life histories. This “individualizing” quickly moves into the social field, drawing therapists into a complex web of social relations. Their need to tell stories is related to their need to decipher these social relations. (Mattingly 1998: 279)

Mattingly provides two examples of this skill at work. The first concerns uncertainty about a diagnosis. A three-year-old boy has been diagnosed as having “sensory integration dysfunction,” but the OT working with him suspects that he might have a head injury. The evidence for the initial diagnosis is that the child is “opposed to any kind of movement and tactilely defensive” (Mattingly 1998: 282–283). To the OT, though, he looks more like a head-injured child because “he also has a lot of things … that sensory integration dysfunction doesn’t have” (283). To try to resolve her puzzlement, the OT decides to talk to the family. She wants to hear a story about what happened to the boy. The mother relates that he was dropped on his head by an uncle, but she does not seem to be upset when she conveys that information, and she does not seem to take the incident very seriously. The mother’s story and her accompanying demeanor make the OT wonder about potential neglect and abuse. The OT continues her investigation by trying to discover what happened at the time of the injury. She learns that the possibility of a head injury was not considered when the accident occurred. She also learns that the boy was brought to the emergency room, spent four hours in the hospital, and then was released with no follow-up. A pediatrician’s subsequent referral to a therapist for rehabilitation mentioned only “developmental delay.” The OT now constructs a story about a patient getting lost in the system. The family

176 

Chapter 7

is poor, does not speak English, and has no insurance. Moreover, she begins to feel culpable because the boy is receiving therapy only infrequently. The family cannot afford more therapy sessions, so she cannot assess the boy’s abilities properly and devise an appropriate treatment plan. She realizes that he needs to be in a school program because therapy through schools is free, but she does not know how to arrange that. The OT is striving to understand an unclear, confusing, distressing situation by constructing stories that will make sense of what has happened, what is wrong with the boy, and how she can help the boy. But she, along with the boy and his family, is thwarted: The therapist found that she was lost in the story; she could not find a place in which she could be effective. The patient was lost as well. He was never properly diagnosed or treated. The family, too, was lost. They were paying for services they could not afford, and no one was helping them locate good resources for their child. (Mattingly 1998: 284)

Despite that dismaying outcome, what the OT did is rational. Mattingly regards her storytelling as part of “a reasoning process” (1998: 284). Her focus on the process is right, as is her recognition of its rationality. Given that formal reason equates rationality with reasoning, though, she is compelled to characterize the OT’s storytelling as not reasoning. It is not, of course, the kind of reasoning that formal reason endorses, and it might be tempting to follow Nelson and describe it as reasoning in a “loose” sense of the word. But it is better to abandon the constraints of formal reason and appreciate the broader conception of rationality that is operating here. The OT uses the four tools of nonformal reason. She carefully observes the behavior of the boy as she is assessing him, and the demeanor of his mother as she explains how her son was injured. She uses formal and nonformal reasoning methods to make inferences about what might have happened to the boy and then critically evaluates the conclusions of those inferences, notably in interpreting the story the boy’s mother tells her. Being dropped by an uncle might have been just an accident, but the mother’s apparent lack of concern, in the judgment of the OT, warrants further investigation. So she entertains the possibility of culpability on the part of the hospital and devises a new process for investigating that alternative. Her creative construction leads to an understanding of how institutional, economic, and cultural factors might have contributed to the boy and his family getting lost in the system, as well as to identification of a possible source of help for the boy. In sum, nonformal reason utilizes storytelling in a rational process for addressing this problem.

Designing Deliberation 

177

Mattingly’s second example illustrates how decisions about what to do can emerge from story telling. Stories can “guide future actions,” in her view, because “the process of telling stories about a patient’s past history or ongoing involvement in treatment is very often a process of exploring and negotiating a vision of the future good” (Mattingly 1998: 286). Telling stories to explore and negotiate “a vision of the future good” is similar to the process of constructing imaginative scenarios that was used by the women who were deciding whether to try to become pregnant after receiving genetic counseling. Mattingly’s example is a collective enterprise, however, because it involves five psychiatric OTs who are trying to understand why a patient has been readmitted to their unit, in the course of which they survey alternative conceptions of the good for the patient and end up with a decision about what is best for him. The patient is a young teenager who got upset about his mother going on a date with a new boyfriend and started clinging to her and pleading with her not to leave. The boyfriend beat him up, after which he threw a brick at the boyfriend’s car and slashed the tires. Then his older brother came home and beat him up. His mother wants him put into foster care, but he wants to stay at home. Because the boy is strongly attached to his mother, the psychiatrists are worried about how a separation might affect him. The OTs point out that he does well when he is on their unit. Much happens, seamlessly and invisibly, in the one-page transcript of the conversation among the five OTs. Blame shifts as the story unfolds, and the boy clearly ends up being the victim. The OTs conclude that the boy should go to a foster home, “not because they think the boy is too much trouble, but because the family does not take care of him” (Mattingly 1998: 288). Their deliberation is brief because the story is vivid and compelling. The OTs feel no need to consider other ways of framing the problem and proceed directly to problem-solving by proposing ways in which they can help him deal with the separation. How does so much happen so quickly and so smoothly? Mattingly’s explanation is that “moral arguments are buried within clinical stories”: the therapists’ interpretive process of telling the story is inextricably intertwined with identifying the morally good treatment for this patient. But the good is never explicitly discussed. Rather, it is naturalized in the very process of the telling itself, the “obvious” result of a recital of apparently relevant facts. But however invisible the reasoning process here, disguised as simply getting the facts straight, the therapists’ discussion belongs much more to an Aristotelian notion of practical judgment than to clinical reasoning as applied science. (1998: 289)

178 

Chapter 7

Mattingly cites two crucial features of Aristotelian practical reasoning. One is that it addresses the “deepest practical question: how should one live?” and the other is that it “understands action as a judgment rather than an application of general rules to a particular case” (1998: 289; emphasis in original). And that judgment emanates from deliberation that is specific and contextual. It is deliberation about particular people with particular histories at a particular time in a particular place. For that reason, Mattingly explains, judging how to act requires capacities to “read” intentions of actors, to imagine outcomes based on a variety of possible actions, to discern what actions will be considered appropriate by others, and to understand the important historical contexts of which any given moment or any single act is just one part. Narratively speaking, practical judgment requires the actor to answer such questions as: What story or stories am I a part of here. How will these stories change if I take this act in this way rather than that one. (1998: 289–290)

Mattingly appreciates that deliberation is uncertain and risky. Indeed, the view of the OTs might not prevail because they do not have the authority and power of the psychiatrists and because the treating psychiatrist has in the past sided with the boy. Mattingly also recognizes that “the process of story construction feels more like ‘getting the facts straight’ than like plotting an argument” (1998: 290). For Mattingly, explicitly “plotting an argument” is an exercise of formal reason. But instead it could be the kind of marshalling of evidence that a lawyer does when summing up in a trial or the way a detective constructs an explanation of how the crime happened. Both cases are rationally argued, certainly, and the weight of evidence is brought to bear, but both proceed from stories that cannot be reduced to simple logical inferences. Moreover, a sharp dichotomy between normative and descriptive premises cannot always be maintained. In stories, characterizations of persons can slip between describing and morally evaluating because learning can involve both, and this flexibility contributes importantly to the power of narratives. Stories and dramatic rehearsals are rationalized essentially, not by formal arguments, but by the rationality that nonformal reason accords them. Cheryl Misak (2005, 2008) provides another illuminating account of narrative ethics, one that merges experience and narratives in ethical deliberation and is “a very practical kind of ethics” (Misak 2008: 615). Her work emerges from her experience as a patient with multiple organ failure in an intensive care unit (ICU) on full life support for a month. Like many patients in ICUs, she was in delirium. Throughout her stay in the ICU, the staff treated her as if she were autonomous and competent to consent to treatment. She believes that she should have been treated paternalistically:

Designing Deliberation 

179

“It is better to follow the policy of doing what is in the patient’s best interests, despite the fact that this will be viewed as highly paternalistic by the patient herself and perhaps by much of the community of health care professionals and ethicists” (615). Advocates of evidence-based medicine and moral theorists would be stunned by her heretical view. But she learned from her experience. Misak describes narratives as “accounts that give coherence or shape to events and are thus freighted with interpretation, motivation, and other dents to what we think of as objectivity” (616). Both narratives and the experiences conveyed by narratives are rational: “Narratives … are reason structured or open to rational criticism, and this allows them to play a legitimate and essential role in ethics,” and “experience is central to rational deliberation” (618). With its focus on experience, Misak’s narrative ethics is naturalistic, just as our ethics is. For her, “real life cases or real life experiences are what moral thinking turns on—it cannot do without them” (632). For her the aim of moral deliberation is “to take moral insight, judgment, and argument on their merits” (626). “If ethics is a subject on which we can legitimately deliberate,” Misak realizes, “then our beliefs must be responsive to further experience, further argument, the contrary reports of others, and so on,” and to do that, we need to use “the full range of our rational scrutiny” (628). There are many affinities between her conception of ethics and our conception of ethics—what ethics is and how ethics is to be done. Although Misak’s view is grounded in rationality, she does not expound either reason or rationality. Nonformal reason might be a welcome complement to her moral deliberation. Although the constructive capacity and the sensitivity of narratives have been emphasized, narratives also can powerfully shape and reinforce biased, erroneous, and ignorant (and even worse) positions.2 Minor mistakes can be amplified to become malevolent malfeasances, heroics can be missed because they are unacknowledged, and so on. Nonformal reason is needed to scrutinize narratives critically for such outcomes, test them to limn their extent, and create alternative contexts and processes that diminish their negative impacts and enhance their positive contributions. That is, nonformal reason is used to judge the impacts of narratives, not just to judge what is in narratives. 7.3  Deliberation Shared The conflict between respecting patient self-determination (applying the principle of autonomy) and protecting patient welfare (applying the principles of nonmaleficence and beneficence) has played out in bioethics as

180 

Chapter 7

a debate about the locus of decision making in health care: should the responsibility for making a decision reside with the patient or the physician? Framing the conflict in that way imposes a dichotomous choice that fractures participation in both decision making about treatment and the resolution of ethical problems. Authority to decide is vested in one of the parties, with the other consequently disenfranchised. That fracture proceeds, in large part, from the presumptively sharp dichotomy between facts and values that we saw in the genetic counseling example in chapter 1. According to the prevailing roles of genetic counselor and patient, the counselor, qua professional and medical scientist, delivers the facts, and the patient provides the values. Likewise with doctor and patient. How can decision making be shared when the question is who decides, and the answer has to be one or the other. Dan Brock (1991) advises the doctor to find a “delicate balancing” between being an advocate for the health and well-being of patients and respecting their self-determination. Doctors have those responsibilities, and they have to make judgments to achieve that “delicate balancing,” but it is not about decision making by doctors and patients. That decision making is not shared. The doctor decides whether to prescribe a drug. The patient decides whether the prescription is filled. Both doctor and patient can participate in discussing and examining a problem and how to handle that problem. They are sharing in a colloquy, and they are engaging in a process of deliberation that is, hopefully, well designed and collaborative. The doctor and patient may agree about what to do, but the decisions to be made are their own separate decisions. 7.3.1  Deliberating with a Doctor In North America, the locus of decision making in medicine has recently shifted. For centuries the expertise of the doctor accorded decision making to the doctor. Physicians possess medical education, practical training, and accumulated experience, and with the emergence of professionalism, their knowledge, skills, and proficiency are tested in licensing exams and must be periodically updated to satisfy continuing education requirements. The medical expertise of physicians has been taken to entitle them to decide what is best for their patients. Recently, though, the scope of that entitlement has shrunk. Doctors, of course, still have expertise in diagnosing medical conditions and identifying appropriate treatments, but they are no longer entitled to decide whether the reasonable outcomes of treatment are worth pursuing, let alone when treatments should not be admitted because they are uncertain or controversial. Decisions about the acceptability of treatment now are determined by the values of patients. The doctor knows

Designing Deliberation 

181

what can be done, and often what should be recommended, but the patient decides what will be done. Medical paternalism has largely been repudiated. Responsibility for decision making has shifted from the physician to the patient, at least in theory if not always in clinical practice. (And, recalling Misak above, not always with the patient’s moral support.) The emergence of respect for patient autonomy is represented in bioethics by models, that is, idealizations, of the physician-patient relationship. Yet the notion of sharing remains popular in discussions of physicianpatient decision making. However, as long as the dichotomy between facts and values prevails and the focus remains on who decides rather than how to decide, what sharing means and how it can be achieved will remain obscure. Aspects of real sharing would be easier to identify were decision making regarded as a process that is a collaborative, mutually respectful dialogue (a) not just about information but also about how widely the situation is to be construed given the circumstances, what values are relevant to the situation (compare here the genetic counseling patients), and what alternative courses of action might be pursued and with what consequences; (b) involving elements such as negotiation and compromise; and (c) wherein the decision-making context encompasses not just making the immediate decision but projecting a cordial, productive relationship into the future. When, for example, physicians balk at prescribing antibiotics that patients request, for themselves or for a child, they might discuss not only why the patient wants the drug but why the physician is unwilling to prescribe it and mutually search for a resolution that is acceptable to both.3 Doctor and patient can jointly participate by contributing proposals and reflectively and critically assessing them, including the values involved; by behaving in a manner that displays respect, concern, and consideration for the other person; and by accepting mutual responsibility for the position they reach. Ezekiel and Linda Emanuel (1992) provide four models of the physician-patient relationship. The first two are familiar. One is the paternalistic model, also known as the parental or priestly model, which assigns decision making to the doctor. The other is the informative model, also called the engineering or consumer model, which ascribes decision making to the patient. The two new models are the interpretive model and the deliberative model. In the interpretive model the task of the physician is to elucidate the patient’s values and help the patient choose the medical options that best realize those values. The physician functions as a counselor, examining the patient’s goals, aspirations, commitments, and character with the patient in an attempt to develop and specify what is vague or inchoate and to reach a coherent overall understanding of the patient’s values. The

182 

Chapter 7

patient decides which values are important, and the physician does not judge those values. The physician then helps the patient understand what medical tests, treatments, and alternatives are consonant with those values; the patient chooses; and the physician implements the choices the patient makes based on the enhanced self-understanding that has been mutually generated. In the deliberative model, the physician helps the patient determine and choose the best “health-related” values that can be realized in a situation. The physician may discuss only “health-related” values, that is, “values that affect or are affected by the patient’s disease and treatments,” and may use no more than moral persuasion in promoting the healthrelated values the physician deems most worthy. In this model the physician is said to act as a friend. Because each of their models incorporates a different account of autonomy—respectively patient assent, control, moral education, and moral self-understanding—Emanuel and Emanuel hold that no single model can be endorsed. They suggest that the models should be assessed, at least in part, with respect to the adequacy of the notions of autonomy they embody. Nevertheless, they claim that “descriptively and prescriptively … the ideal physician-patient relationship is the deliberative model” (1992: 2225), adding the contextual qualification that different models may be appropriate in different clinical situations. Both the interpretive and the deliberative models have serious problems, though. In the interpretive model, values remain the exclusive domain of the patient, and basic values are never challenged by the physician. Consequently, it is in an important sense a nondevelopmental model, because whatever improved self-understanding results is confined to clarification and not transformation—unless, as in the case of the women who, after receiving genetic counseling, determined for themselves how they would decide whether they would try to get pregnant, the enhanced self-understanding is initiated by the patient. The model posits only a weak interpretive relationship, because there is no improvement in the patient’s understanding of how to enhance their suite of values or how to assign priorities to its constitutive values in varying and new contexts. A stronger model would give the relationship a developmental (i.e., maturity-gaining) aim by including mutual exploration of the values of the patient to support the patient and to strengthen the patient’s independent critical capacities. The deliberative model has two serious difficulties. One is the vagueness of the notion of “health-related” values. With more serious illnesses and chronic illnesses especially, virtually all of a patient’s values could be affected, at least in the sense of being diminished or compromised.

Designing Deliberation 

183

Diminished or not, when profound consequences for quality of life loom, many general values, for example, relieving suffering, retaining autonomy, preserving dignity, and fostering courage, will unavoidably be in play; these reach across all the dimensions of life, including health. Given such complexity and contextuality, let alone the hazy notion of health, how can a line between health-related values and non-health-related values be drawn? Without such a line the concept of health-related values cannot circumscribe the ambit of physicians’ decision making, particularly where that line is most needed. This difficulty exacerbates the other difficulty. Emanuel and Emanuel recognize that even though the deliberative physician may not go “… beyond persuasion, in practice the deliberative physician may occasionally lapse into paternalism” (1992: 2225). This attempt to design decision-making models for physicians is trapped in a dilemma of its own making: if a sharp fact/value distinction is accepted, participation in moral deliberation and decision making is fractured, and if the distinction is rejected, paternalism threatens and, moreover, bereft of the protection that health-related values are supposed to afford patients. All four models also impose substantive obligations on doctors. In the paternalistic model doctors are obliged to put the patient’s interests above their own and to seek help when their knowledge is insufficient. In the informative model physicians have the obligations to provide truthful information, to maintain their competence, and to consult with others when their knowledge or skills are insufficient. In the interpretive model physicians additionally have the obligation to take the patient through a process of self-clarification. The obligations in the deliberative model are those of a counselor, teacher, or friend. With respect to a counselor or teacher, the obligations are to inform, to clarify, and to instruct, that is, to help the patient understand and mature, but not to go beyond respectful persuasion. With respect to a friend, the obligations are to be interested, concerned, supportive, and loyal, and perhaps also include the obligations of counselor or teacher, although friendship seems to require them less centrally, and they may be in tension with being supportive and loyal, that is, with doing what is needed to preserve the friendship. For Emanuel and Emanuel, these obligations are relative to their models, but the expansive, variegated morality of professionals and patients or clients cannot be decocted to the obligations attached to the models. Not only are all the models partial, the adoption of any model depends on higher-order value judgments that transcend the contexts of the models. In addition, value judgments are needed within the models. In the favored deliberative model, for example, determining the relevant health values

184 

Chapter 7

in the circumstances and drawing the boundaries of respectful persuasion are matters of judgment. The role established by a model can include only selected aspects of a sweeping, open process of moral decision making, where the principal challenge remains how to properly design participation in it. Fortunately, nonformal reason can support extensive, mature developmental relationships in which locating boundaries is part of what is learned and negotiated. 7.3.2  Deliberating with a Clinical Ethicist In his book Troubled Voices, Richard Zaner recounts a number of situations in which he, a philosopher trained in phenomenology and working in a hospital, provided clinical ethics consultations. At the end of the book, Zaner reflects about what these situations have in common: Each of them confronts the people involved—and ourselves, to the extent that we can feel the difficult issues these people faced—with a basic challenge to their sense of self. What and who we are, what we hope to be and become, even whether we will continue to be at all, is in one way or another at stake in these circumstances. This poses basic moral questions. (Zaner 1993: 137)

That richer conception of morality—a morality that is about whole persons and personal development, not just principles, about what a person can hope and strive to become, not just what are permissible and obligatory—is the morality of the women making post-counseling reproductive decisions, of K’aila’s mother, and also of Dr. P. But it is not the morality of a bioethics in the guise of applied ethics. Zaner wonders about the place of ethics in his consultations, one of which involves the Olands. The wife and mother in this family wants to have her life supports withdrawn, but the husband and father wants “everything possible” done. The matter is resolved only when, in discussions with Zaner, the husband’s grief and guilt is expressed, allowing him to respond to his wife’s circumstances empathetically.4 Zaner’s problem, like so many real bioethics problems, is a moral problem, but not the kind of pristine moral problem that appears in bioethics textbooks, purged of detailed circumstances and messy contexts to make way for moral theory. Clinical ethicists who equate morality with a theory can lament that one of their consultations was about a communication problem, not any moral problem. As Zaner’s consult reveals, poor communication can be an important ingredient of a complex moral problem, as can lack of trust and lack of understanding. His consult does not invoke moral principles or a moral theory. Espousing autonomy is easy; helping people make difficult decisions

Designing Deliberation 

185

in a coherent, sensitive, rational manner is hard. Zaner explains that the practitioner has to examine clinical encounters thoroughly “before making up one’s mind about such austere matters as ethical principles” (1994: 211). Musing on the Oland case, Zaner asks, “What did anything that went on have to do with ‘ethics’—a question I’ve grown used to asking myself with embarrassing regularity” (1993: 32). Then he describes his job as a clinical ethicist: All I did was convene a meeting, moderate it, and then back out to let things take their course. I had been asked to consult by the attending physician. As he told me later, he thought that the family might feel somewhat freer to talk about things with “people like you.” When he first talked with me, interestingly, he too felt quite strongly that something still unspoken was troubling this family. In asking me for a consult, he hoped that I might be able to ferret it out. He also reminded me of my own words in an earlier conversation with him when I told him that I often felt rather like a detective, and he proceeded to lay out what he took to be a “clue.” When he had mentioned the possibility of a feeding tube, he had noted a kind of strained reaction. The atmosphere markedly changed, and he decided that, as this was not an immediate issue, he would be better off not pursuing the topic just then. (Zaner 1993: 33)

That “clue” does lead, albeit in an unexpected way, to a resolution of the family’s troubles. But is it a moral problem, or a communication problem that perhaps would be handled better by a social worker? Zaner is brought in to deal with what looks like a prototypical bioethics problem—the withdrawal of life-sustaining treatment—but he does not wield the principles of applied bioethics. So is this one of those consultations that ends up disappointing clinical ethicists itching to do “real ethics.” Zaner explains how he understands “real ethics”: This encounter happened many years after I had begun to get involved in these sorts of situations and had already learned a good deal about them, enough to know right off that ethical issues are most often deeply buried, rarely recognized as such, and may require a good deal of indirectness to deal with them appropriately. Kierkegaard, that master of irony and the subtleties of human life, knew this best of all, for he knew enough to know that such matters can only be talked about by not talking about them directly. Indirect discourse, in his words, is the only way to go about understanding things human, especially matters of ethics. So, what I had learned to do is never to talk about “ethical issues,” or rather to talk about them by never talking about them. What happened in the case of the Olands seems to fit that idea rather well: feeling that something was really bothering him [the husband and father], it seemed to me that the thing to do was just give him a chance to talk, to come to what was wrong in his own way and at his

186 

Chapter 7

own pace. Even when he began to talk, though, the thing lying deeply buried beneath even his terrible grief still took time to emerge, and then did so only very gradually, as if testing the light of day for any treacherous snares that might be lying about. Not talking about his grief and guilt directly was a way of letting him talk about them indirectly. People faced with such difficult decisions know this well. Such is the irony of human experience communicated. (Zaner 1993: 34–35)

Applied bioethics is deeply rooted in the logic-based tradition of analytic philosophy, the main Western alternative to the phenomenological tradition in which Zaner was trained. Irony, subtlety, and indirection are anathema to analytic philosophy, which prizes clarity, precision, rigor, and systematicity in all things. (These latter values have their proper roles; it is the framework within which they are invoked that makes all the difference.) Was Zaner wrong not to remind the family that Mrs. Oland was competent and that her autonomy ought to be respected? That would have been the open, direct, morally prescribed way to respond according to an applied bioethics conception of the Olands’ situation. But does that conception of ethics appreciate the Olands’ plight? Did they have a moral problem, and if so, what, other than engaging the principles of applied bioethics, would make it reasonable to regard their conflict as “moral”? Zaner found a way to reveal the momentous responses called forth by grief and guilt and then to integrate them into a healing, communicating moral integrity for the Olands. These are, as Zaner says, basic ethical issues. If he is right that ethical issues are most often deeply buried and rarely recognized as ethical issues and that the lesson of human experience is that they are best handled indirectly, applied bioethics badly misses and distorts much of morality. By Zaner’s own account, his intervention was effective, but Zaner’s effectiveness is not the effectiveness of applied bioethics. What, then, is clinical ethics? How do clinical encounters go? How do conversations illuminate and elucidate? First and foremost, Zaner listens, which he calls “letting people be” (Zaner 2004: 15). Patients want to do the right thing, but their beliefs and values do not provide the clarity and direction they need. So, as Arthur Frank puts it, “they need someone to listen to them while they talk their way into new beliefs, founded on the old ones, but supplementing and revising them considerably” (Frank 2011: 157). Zaner solicits and refines stories. He is a kind, patient, sympathetic listener imbued with a philosopher’s critical acumen and passion for clarity. Patients have to tell and retell stories about themselves to see where the stories go and whether the endings seem right to them. In this conversation there is a disciplined, twofold creative construction, carried out by the patients, clarifying for themselves who they are, and by Zaner, clarifying

Designing Deliberation 

187

to himself the patients’ situations and problems, both creations honed by sharply crafted critical appraisals. Abstract values are of little or no help here. They have to become what Frank calls “working values” (Frank 2011: 159), and values are working when they are situated in and exemplified by the particularity and contextuality of stories. The ending of a story seems right enough when it fits, that is, when it makes sense and feels right. And “sometimes,” Zaner reports, “it takes a lot of telling before you get it right” (Zaner 2004: 9). What the women who received genetic counseling do in their imagination, Zaner’s patients do in their words. For both, the process is crucial to their deliberation. Doing what is right is getting the process right to get the story right. The story is the patient’s story, and the judgment that the story has it right is the patient’s judgment. Zaner shares in the process by helping the story to unfold and by being what Frank calls a “witness” to the telling of the story. The job of philosopher Zaner as clinical ethicist is to listen to their stories with understanding and to get his patients to tell their stories until they get them right. He is their Socratic narrative midwife, prompting them to examine their lives and using the resources of nonformal reason to render their narrative deliberation rational. Uncertainty about what constitutes a moral problem poses a practical challenge for ethics consultants and ethics committees (and a profound challenge for the field of bioethics, given its ambition to be recognized as a profession). When an ethics consultant is appointed or an ethics committee is formed in a health care institution, procedures governing the work of the consultant or the committee have to be established. Those procedures must specify not only how work will be done but what kind of work will be done. One tricky issue, politically as well as practically, is who may have access to the consultant or the committee. An even more tricky issue is what problems may be brought to the consultant or the committee. The solution to the latter issue seems obvious: an ethics consultant or committee will consider only moral problems. But what, then, is a moral problem? Ethics consultants lament that many of their consultations end up involving communication problems rather than moral problems. They seem distressed that they are wasting their time on minor, mundane problems that should be handled by somebody else—a social worker, probably—instead of confronting the serious moral issues they were trained to resolve. How do they make that discrimination? Why is one problem merely a communications problem and another a genuine moral problem? Can distinct moral problems really be identified, isolated, and extracted, intact and whole, from the complex situations, relationships, and contexts in which they are

188 

Chapter 7

deeply embedded? Is a clinical ethicist, as applied ethics suggests, a logician deriving the correct solution to a well defined problem, or, as Zaner suggests, a detective trying to figure out what the particular problem of particular people is? The rationality of the clinical ethicist as logician is the narrow conception of formal reason. The rationality of the clinical ethicist as detective is the broad conception of nonformal reason. 7.4  Institutionalized Deliberation Institutional design can fracture the moral decision making of professionals. Renée Anspach’s study of two intensive care nurseries illustrates that rupture between doctors and nurses. Her research proceeds from the appreciation that medical decision making is a social process, one aspect of which is how organizations and settings “shape and structure” decisions about diagnosis and treatment (Anspach 1987: 215). Anspach found that disagreements between the doctors and nurses in these units about whether to treat infants aggressively did not proceed from conflicts about the governing moral principles but conflicts about the prognoses for the infants. Even when doctors and nurses agreed about the relevant principle in a lifeand-death situation, they often could not agree about the likely prognosis for an infant. This rift intrudes on ethical decision making for the infants every bit as much as does ethical disagreement. How did these disagreements arise? From, Anspach discovered, the different kinds of knowledge on which doctors and nurses base their prognostic judgments.5 And these different kinds of knowledge result from the different kinds of work doctors and nurses do.6 The prognostic judgments of doctors proceed from a combination of what Anspach calls “technological cues” and “perceptual cues.” Technological cues are the information obtained from diagnostic testing. Perceptual cues are the information obtained by examining the patient, using techniques such as palpation, percussion, and, most commonly, passive observation. The prognostic judgments of nurses proceed from what Anspach calls “interactive cues.” Interactive cues are the information generated by social interactions between the patient and the health care practitioner. With adults, much of that information comes from a clinical history, but with infants and children, it also comes from nonverbal aspects of communication. The types of information that doctors and nurses value and use depend on the nature of their professional training and the nature of their jobs. Doctors regularly but periodically come to the intensive care nursery to review the results of tests, read charts, examine the babies, and conduct rounds.

Designing Deliberation 

189

They spend relatively little concentrated time with the babies. Moreover, doctors change. Attending physicians have rotating shifts in the unit, and, as part of their training, residents rotate through the unit to assignments in other units. In contrast, nurses have extended, continuous assignments to patients, and they work long shifts in the unit, during which they are caring for only a few babies. They touch them, hold them, and comfort them. They see how alert they are, how much they respond to stimuli such as sounds, light, and touch, how often they cry, how “floppy” they are. Over the time they spend with their patients, they detect patterns and changes. Anspach points out that the prognostic judgments of doctors and nurses, and the decisions based on them, emanate from “the standpoint of a very partial knowledge base, selectively filtered through the lens of structural circumstances” (1987: 227). The knowledge of the doctors is skewed toward the generalities they have been taught, learned from experience, and read in medical journals. The knowledge of the nurses is skewed toward the particularities of what they have observed in their interactions with the babies, for example, their energy levels, alertness, discomfort, and “will to live.” When compared to that of the physicians, the nurses’ knowledge can be easily discounted or dismissed as idiosyncratic, biased impressions that are less relevant and reliable than the generic lessons of established medical practices. Yet the judgments of the nurses are not just subjective hunches or, in the parlance of philosophy, intuitions. One unit nurse’s account of their knowledge provides a more realistic, more charitable assessment: I think that with a lot of nurses that have been here for a while, the gut feeling has developed from a lot of observation that they have recorded subconsciously and never put on paper, and that no one’s ever written about and I think that there are a lot of those things, so when people say, “I don’t like that kid, there’s something wrong with him,” the gut feeling’s really based on a lot of informal study that they don’t notice that they’re doing. (Anspach 1987: 222–223)

From this perspective the views that nurses develop from their accumulated experience look much less like a “gut feeling” (or an intuition) and much more like a rational judgment that proceeds from the use of the four resources of nonformal reason. This construal of their knowledge offers the prospect of a more reasonable, more ethical response to managing conflicts between doctors and nurses. There is, however, a further complication with the clash between what Anspach calls the physicians’ perspective of detachment and the nurses’ perspective of engagement, the former based on conclusions derived from diagnostic technology, physical findings, and epidemiological studies, and

190 

Chapter 7

the latter based on inferences from extended contact and (nonverbal) social interaction with the babies. Within the culture of academic, tertiary care medicine, which is where intensive care nurseries reside, and within the history of medicine, the interactive cues perceived by the nurses and the knowledge those cues impart are, as Anspach recognizes, “devalued data” (1987: 229). The nurses’ knowledge might also be devalued because most nurses are women and the knowledge of women has historically been dismissed or stereotyped as emotional, quirky, and superficial. The fracturing of knowledge and of participation in decision making that Anspach’s study reveals is a serious problem that has many ramifying institutional dimensions. Anspach appreciates that the organization of work in an intensive care nursery, within the background culture of health care, can have a deleterious impact on participation in decision making: “Those who have the most patient contact (the nurses) and the most at stake (the parents) have the least authority in life-and-death decisions” (Anspach 1987: 230). She also appreciates that the fracturing of both knowledge and participation must be removed if the quality of the decision making is to be improved: To the extent that decisions cannot be extricated from the social organization of the intensive care nursery, broader changes in that organization may be necessary. A culture which permits the exploration of other modes of knowledge and a social structure which facilitates greater interaction among physicians, patients, and parents, may contribute to more informed and equitable decision-making. (Anspach 1987: 230)

And, we add, could contribute decisively to genuine role accountability on everyone’s part. Although the problem of fractured participation is not exclusively a matter of institutional design, it is largely created and perpetuated by institutional design. Yet institutional design dedicated to enhanced moral decision making that goes beyond hiring a clinical ethicist and establishing an ethics committee is rare, and constitutes a daunting enterprise. Designing processes that, on the one hand, facilitate more genuinely collaborative, deliberative, and creative decision making than what prevails in the practices of ward rounds, unit rounds, teaching rounds, grand rounds, and ethics rounds, and, on the other hand, are compatible with the physician’s ultimate authority and responsibility for a decision is an intimidating challenge. Likewise trying to design ways of promoting better communication and improving participation in decision making given work schedules, shift changes, and staff rotations; or trying to design the educations of health

Designing Deliberation 

191

care professionals in a way that fosters more informed, complementary, and novel interdisciplinary decision making. Despite such formidable challenges, institutional design is the route to more participatory, informed, and equitable decision making. Individuals often feel alienated from the institutions with which they must deal as employees, patients, and clients, and alienation caused by features such as the unintelligibility of processes, powerlessness to affect processes, the distance between inputs and outcomes, and the lack of adaptability to local conditions has been spreading as the size, concentration, and complexity of Western institutions, both public and private, have increased. The only positive alternative to blind destruction and backward-looking rural utopian options is to improve participatory design. Properly designed participation substantially reduces the alienation of individuals by helping to reunite reason and motivation. Doing that is inherently ethical and also forms solid ground for the ethics of professional relationships, with patients and clients and among professionals.7 7.5  Conclusion Matters of moment in life are matters for deliberation. To be meaningful and helpful, however, that deliberation must have an appropriate design, an appropriate overall shape and dynamics. This requirement does not impose a single, theoretical design; for example, it cannot be an abstract calculation of likely advantages and disadvantages. Rather, it must be a particular and contextual examination in which critical assessment comes not just from our intellect but from our entire being. The imaginative scenarios of dramatic rehearsal and the stories of narrative ethics are ways of deliberating relevantly and, with nonformal reason, rationally about our lives. The preceding cases emphasize the great diversity and richness of deliberative processes. Together these processes contain the whole expression of nonformal rationality in all its internal complicatedness and its contextdependent variety, including its constant improvement. Too complicated, self-transforming, and context dependent to be reduced to any simple characterization, much less any detailed formula, deliberating is best understood through examples like those above and others found throughout the book: the women deliberating after genetic counseling in chapter 1, the vignettes in chapters 2 and 5, the ICU staff coming to terms with the encephalitis case in chapter 6, and Rapoport’s debates in chapter 8.8 This richness is no unstructured chaos; there are powerful organizing and ordering forces at work, principally the necessity that each deliberation satisfy

192 

Chapter 7

the requirements of being nonformally rational, thus applying the four resources and utilizing compromise and wide reflective equilibrium. Moreover, the construction of a deliberation is a problem-solving activity, bringing to bear the structure of the problem-solving process (section 4.3 above) to help organize the construction of rational deliberation. This process, recall, has at its heart a learning process from which issues both particular design solutions and a general design expertise manifested in the case studies above.9 Deliberation is most reliable and productive when it is a collective, social enterprise. Models of the physician-patient relationship are attempts to design joint decision making, but they focus on who is entitled to be the decision maker, not on the process of deliberation. The goal of these models is not surprising because, as we saw in the presumption about how the women who received genetic counseling should make their decisions, the models likewise assume that there is a sharp dichotomy between facts and values and that rationality is reasoning from principles. Decision making can be shared only when a genuinely deliberative process is shared, as revealed in the storytelling of the occupational therapists and in Zaner’s approach to ethics consultations, and then decision making is much richer and more complex than what the models of the physician-patient relationship depict. Deliberation can fail or go awry, however, when it occurs within institutional settings and constraints that do not provide the necessary opportunities and resources, and it will not occur at all when the structure or the ethos of an institution obstructs it. The only remedy for those obstacles is institutional redesign.

8  Designing Practices, Institutions, and Processes

C

Much of morality is socially and culturally embedded. Practices and institutions embody recognized values and ideals of the societies and cultures within which they emerge, and more profoundly they are formed and shaped by background assumptions, values, and ideals so deep and expansive that they are invisible.1 The morality that infuses social acceptance of and participation in fundamental institutions such as school, law, health care, social media, entertainment, politics, and religious observance, and that supports diverse practices such as donating time and money to charities, respecting queues, and attending athletic events, influences how we behave, how we live, and who we become. Practices and institutions also create moral problems. Scarce health care resources raise social issues about fairness in allocation (say, of kidneys for transplantation), and personal issues about what to do when resources are not accessible (leading to responses like medical tourism for in vitro fertilization). Socially generated moral problems sometimes impose double binds that force distressing choices between two harms; the only way to escape that intractable bind is by changing or eliminating its source. Because practices and institutions pervade our lives, and because morality pervades practices and institutions, morality is in large part a matter of recognizing how moral problems emanate from the design of practices and institutions and how processes for resolving moral problems are designed. This chapter depicts those core elements of ethics as design and shows how nonformal reason is needed to account for the rationality of that design. 8.1  Fights, Games, and Debates Anatol Rapoport’s classic text Fights, Games, and Debates (1960), is about conflict, a topic that, according to two pioneering game theorists, R. D. Luce and H. Raiffa, preoccupies human beings more than any other, with

D

194 

Chapter 8

only the exceptions of God and love.2 Rapoport begins his examination by providing prototypes of three ways of resolving conflicts. His example of a fight is the verbal and physical combat between Tom Sawyer and a stranger he encounters as he is strolling down a street practicing a new way to whistle. His example of a game is a chess match between Frank Marshall, the United States champion, and Capablanca, the great Cuban player, in the 1918 New York Tournament. His example of a debate is the argument between Settembrini and Naphtha for the soul of a young man in Thomas Mann’s novel The Magic Mountain. In fights, games, and debates, there are at least two participants, which can be individuals or groups—clubs, tribes, and nations fight, play games, and debate, sometimes en masse, sometimes through representatives. Fights, games, and debates represent three competitive modes of interaction with those around us. In these conflicts each participant has at least one objective, and the participants mutually interact to achieve their objectives in a manner that typically involves some combination of cooperation and opposition. Fights are essentially oppositional, the objective being to remove the other participant from the interaction. They often proceed according to explicit or tacit rules, as, for example, in fencing duels and boxing matches, and removal may entail anything from a courteous bow to death. Games range from being wholly cooperative, through “mixed motive,” to wholly oppositional, and the objective of playing the game is to maximize one’s payoff (utility). Even when games are entirely oppositional, however, cooperation in following the rules of the game is required, and, because the challenge of winning the game is what makes playing the game worthwhile, the participants also must cooperate in doing their best to win. Debates are essentially oppositional, the objective being to win the approval of judges or an audience when the arguments are directed to them or to convince the opponent when the participants’ arguments are directed to each other. In the terminology of game theory, fights, purely oppositional games, and debates are all zero-sum games, which means that what one participant wins is lost to all the other participants. Cooperative games are sum-increasing because by cooperating all participants increase their payoffs and hence the total payoff for the game. Fights and debates, while essentially oppositional, can be embedded in an overarching cooperative process where all participants increase their payoffs, as when a fight is staged for entertainment or a debate is used as a respectful, collaborative way to work through an issue, with primacy given to the quality of the exploration. Before looking more closely at these three forms of conflict, it is worthwhile to note Rapoport’s summary of the contest between Marshall and

Designing Practices, Institutions, and Processes 

195

Capablanca because of how compactly it illustrates the interaction of formal and nonformal reason. He begins with an analysis of what Marshall was doing years before their match, when he was becoming even more skilled by playing against himself, an endeavor of which Rapoport suggests we could aptly both “call it practicing” and “call it research”: Until about fifteen years ago it did not occur to many people to associate research with games. Yet the thoughts of the mathematician and of the chess player are of the same sort. Both move in mazes of deduction. Stripped to essentials, the “skeleton” of these thoughts is of this nature: “If so, then so, unless so, in which case so.… But on the other hand, if so or so, then this, which leads to that …” Put different kinds of “flesh” on these bones, and you have mathematics or logic or chess or bridge or the “elementary” reasoning of Sherlock Holmes, or the scheming of a diplomat.” (Rapoport 1960: 2)

Formal deductive reasoning is the skeleton of thinking and indispensable, but it needs not just flesh but supplementation by judgment. At that time Marshall was using mazes of deduction to investigate defenses for Black to the powerful Ruy Lopez game of White, against which he hated to play. One night Marshall discovered moves that set subtle traps for White, and when he checked, these new variations were not in any of the Ruy Lopez games he could find. Marshall’s quandary was when to test his discovery because if it did work in a game, it would work only once. He avoided playing the Ruy Lopez game for ten years, waiting for a proper opportunity. Capablanca finally was the deserving opponent and the worthy cooperator. Having studied all of Marshall’s games, Capablanca knew that he had not accepted a Ruy Lopez game for a long time. So when Marshall surprisingly offered the Ruy Lopez game to him, realizing that this probably was one of the most important games in Marshall’s career Capablanca could surmise that he “must have investigated as much of the maze of deduction bearing on this variation as was humanly possible” (Rapoport 1960: 4). Capablanca explains his decision to accept Marshall’s invitation: I thought for a while … knowing that I would be subjected to a terrific attack, all the lines of which would of necessity be familiar to my adversary. The lust of battle, however, had been aroused within me. I felt that my judgment and skill were being challenged by a player who had every reason to fear both (as shown by the records of our previous encounters), but who wanted to take advantage of the element of surprise and of the fact of my being unfamiliar with a thing to which he had devoted many nights of toil. … I considered the position then and decided that I was in honor bound, so to speak, to take the pawn and accept the challenge, as my knowledge and judgment told me that my position would be defendable. (Quoted in Rapoport 1960: 4–5)

196 

Chapter 8

Capablanca avoided two traps set by Marshall, then switched from defender to attacker. When it became clear that checkmate was inevitable, Marshall resigned. Rapoport asks, “How does one reduce this drama to logical analysis?” (1960: 5). His answer is: “The game theorists have undertaken this task and in their researches have gone as far beyond what had been known about the logic of strategy as modern logicians have gone beyond the simple syllogisms of Aristotle” (5). The advances of logic and game theory are indeed impressive and important contributions to formal reason. But why must this drama be reduced to the logical analysis of formal reason? This drama is about finite, fallible human beings with feelings and values dealing with uncertainty and taking risks. Marshall had to make a judgment about when, where, and against whom to test his discovery. Capablanca had to make a judgment about whether to play an unknown variant of the Ruy Lopez game. How can those judgments, one of which led to victory and the other to defeat, be made rationally, regardless of their outcomes? Not in the logical terms of formal reason. Rather, by using the tools of nonformal reason. Poring over all those past games involves observation and comprehensive critical assessment focused on analyzing the games and testing strategies by generating mazes of deduction. Nonformal reasoning methods also are utilized, for example, in Capablanca’s assessment of the dangerousness of the traps set for him and the degree to which he can rely on his history of victories over Marshall. Devising a new variation of the Ruy Lopez game is creative construction. The drama proceeds from and inheres in the judgments that had to be made. Trying to reduce it to the logical analysis of formal reason ignores those judgments and thereby destroys the drama. Nonformal reason recognizes and respects those considered judgments as rational, thereby also preserving and conveying the drama. Participants in games and debates have to agree to an institutional structure that defines and governs their interactions. For a game to be well defined, there must be recognized participants—whether persons, companies, political parties, or others—and recognized action options for each participant; and each game-complete bundle of choices—one for each participant—must jointly lead to recognized, unique payoffs for each participant. These requirements must be stable throughout a game for utility maximization to be properly defined. And everyone involved in a game must cooperate socially to produce a stable game. Debates are similar to games, but with even tighter rules that restrict participant behavior to noncoercive, persuasive communication with turn-taking. In typical institutionalized versions of debates, communication is strictly sequenced and

Designing Practices, Institutions, and Processes 

197

timed, but the idea of debate can be generalized to encompass civil discussions of issues generally, discussions extended throughout formal and informal fora across time and space. Fights, by contrast, may range all the way from ritual combat that, like the others, respects agreed rules (dueling, for instance) to warfare that even today mostly breaks any rules to secure power advantage. For fights the primary medium of interaction is power, whether physical or psychological; for games it is utility, whether realized as a currency or as direct goods and services (economic or sociocultural); and for debates it is persuasion, whether directed toward changing beliefs, attitudes, values, principles, or behavior. The ranges of aspects open to change during an interaction are greatest in debates and least in fights. Debates permit all dimensions of a situation to be discussed, explored, and negotiated: accepted facts, values (utilities), institutional arrangements, participants, options, payoffs, and so on. For this reason debates in principle have access to the largest range of outcomes, everything from a simple winlose to overturning the very premises of the original issue in a way that opens up hitherto uncontemplated options. They are confined only by the commitment to civil persuasion, and in the best of them—the fully rational debates—by mutual acceptance of the primacy of the quality of the joint exploration. In the parlance of chapter 7, fully rational debates are a structured form of deliberation. In practice they may be more or less intelligently pursued, but in either case they will not be rationalized by applying any narrow notion of rule-following or maximization since that cannot comprehend strategic search decisions, deep revision of assumptions and content, and the like. By comparison, games are confined by their rules and by the maximization criteria of rationality; games cannot by definition overturn their own constitution, as debates can, or vary the objective of seeking the payoff with the maximum utility. Games must be played out as constituted, with the only room for maneuver being that allowed within their defined strategic options. Changing the constitution of a game may itself constitute a different game that is equally confined in its terms—or may be the objective of a debate or fight—but it is not playing the original game. Negotiation in a game, if any be permitted by its constitution, is confined to the game’s outcome pay-offs, and its usefulness lies in the wide range of situations where maximizing utility in well-structured ways is the predominant action available. (Modern governments try to confine their financial institutions to playing well-structured, stable games that operate under agreed rules, but this goal is regularly undermined by exploitation of loopholes, ambiguities,

198 

Chapter 8

and opaqueness, often driven by perverse incentives—in short, through poor ethical institutional design.) Fights seem as unconfined in method as debates since, as noted, the adversaries can resort to any means useful to gaining power. However, power is a narrow medium in which to communicate, often blocking any real access even to rational self-interest, let alone to subtler questions about the presuppositions of the dispute or opportunities for reformulating the dispute. The mixture of distrust, negative emotions (often visceral), and concentration on assembling and wielding power usually ensures that rigidity. Thus the actual range of accessible outcomes is typically very narrow for fights. Actual conflict resolution processes at whatever societal scale, from those addressed to the more limited stresses and disputes of families to those addressed to the more sweeping negotiations and wars of nations, can be, and often are, mixes of all three kinds of interaction. Dealing effectively with human interactions requires understanding their respective dynamics and possibilities and the dynamics of how issues shift among these modes. Here we wish only to make two key points. First, institutions play key roles in characterizing and stabilizing these different modes of interaction. This much is obvious from the preceding discussion and, more importantly, makes it clear that it is only through the contributions of the embedding civil institutions that we are able to take advantage of the greater richness of games and debates. Precisely because games and debates institutionalize their interactions more thoroughly and stably than fights do, they can more readily be designed to open up new, wider possibilities for social interaction. These possibilities represent the whole weight of societal achievement that intelligence can deliver. Whether prospects of societal advancement can be attained, however, will often depend upon how much societal weight intelligence carries. In any event, without these creative possibilities for improvement we are reduced to devising intelligent methods of mutual destruction, and life is rendered poor, nasty, brutish, and short. Institutions are as pervasive and formative in conflict resolution as they are in setting problem-solving capacities in science—where the field of cutting-edge problems is formed by institutionalized disciplinary interactions in laboratories, journals, conferences, and so on—as well as in determining the ways that problems and the terms of their resolution can be discussed and framed. These institutions and their economic host institutions, for example, universities and commercial companies, supply the funding for research and facilitate its practical application. Institutions shape problem

Designing Practices, Institutions, and Processes 

199

formulation and developmental and conflict-resolution processes, constrain possibilities, and allocate both responsibilities for responding to needs (via roles) and resources to tackle problems. However, institutions are themselves fallible and improvable. They often require external diagnosis and treatment to resolve problems—for example, internal role conflicts— and that need extends from individual businesses to the governing institutions of nations or the planet—for example, the United Nations, the World Bank, and the International Monetary Fund. Institutions themselves must reflexively be the object of rational, ethical improvement as they facilitate the problem solving that itself brings improvement. Second, richer institutionalized modes of interaction are not only more civilized, they are more ethical. Transforming a fight into a game promotes rationality over emotion and limits damage to categories that can be agreed to by the participants in the wider embedding institutions, such as the marketplace or, for politics, the whole society. These outcomes are improvements in multiple broad dimensions of ethics, for instance, procedural justice (e.g., employment benefits), individual rights (e.g., freedom of information), and the reduction of harms (e.g., pollution). In addition, institutional alternatives provide opportunities to argue for respecting various values and ethical principles as constraints on acceptable outcomes and strategies, for instance, in worker occupational health and safety—in short, to advocate for an evolution of game forms as essential to the evolution of a more ethical culture. It is in this role that debates also shine. We can already see in debates the prospect of tackling not only the facts and rights pertinent to a dispute as they initially appear, but also the fears, mistrust, misunderstanding, egocentricity, laziness, and other divisive obstacles that underlie many disputes. Rapoport invites us to expand and enrich processes to foster situations where mutual understanding across emotional and/or cultural divides can flourish and intelligent compromise or, better, mutually transforming cooperation can emerge. To flourish, these creatively benign interactions require institutional settings that deliver mutual trust in the parties’ motivation and security of action; mutually recognizable ways to communicate about values, to express appreciation of others, to own relevant personal feelings, whether helpful and not, and to communicate empathy; willingness to cooperate in the search for agreed resolutions, and so on. In Sydney a young Muslim woman was dismissed as a day-care worker for young children over a dispute about whether she could take time to pray. Furious that her request to perform a short, simple religious duty was met initially with silence by the management, then with dismissal,

200 

Chapter 8

she threatened a lawsuit for discrimination. Fortunately an experienced negotiator was called in. He was soon able to help each side understand how the other felt threatened and had experienced unacceptable behavior, yet despite all that retained the good will necessary to reach an agreed solution, as well as what mutual responsibilities might reasonably be reflected in solution guidelines. In the end, a simple solution was found (she stepped into a vacated quiet office for prayers while others covered child care) that satisfied everyone and cemented better communication and cooperation among staff and with management. This conflict could have played out as a legal game, likely with less harmonious and certainly with more costly results. That alternative would have been far better than resorting to fighting, but transforming the conflict into a liberating debate produced a still better ethical outcome, especially as a model for the children, one marked by mutual understanding and respect and by participatory, creative problem solving, and that introduced a paradigm for future development. The ethical qualities displayed in this example of a debate have already been explored in some detail. In particular, the resolution of the disagreement between the nurse and physician about continuing aggressive treatment for a patient in the intensive care unit (in chapter 7) shows how the interaction of an open debate can achieve a high-quality ethical compromise. The problem there is not just about appropriate treatment, but also about how the parties to the disagreement can respect one another and take the other person’s views seriously enough to warrant a compromise and to explore what would be the most effective, acceptable ethical compromise. (See also the problem for K’aila’s parents, chapter 2.) As these workplace examples show, everyone has some capacity to create institutional arrangements that enlarge the possibilities for ethical conflict resolution. That capacity is of course constrained by a person’s institutional role(s): everyone can work at this within their family, but young children will have less impact than their parents; and doctors and nurses can work at modifying processes for treatment decisions, but not, in everyday circumstances, a nation’s approach to health care policy. Nonetheless, to innovate as we can in institutional contexts is ethically required and the very foundation of a healthy democratic culture. In all cultures constructive contributions are tragically rarer than they could be, limited as they are by, for example, family ties or assumptions about reduced responsibility produced by racism or class, or institutions such as a club or market; but surmounting those limits vastly improves the quality—including the moral quality—of civilized life.

Designing Practices, Institutions, and Processes 

201

Not only does the progression from fights to games to debates reinforce the ubiquitous role of institutions in the constitution of societal ethics, it locates the social source of the institutional creativity that promotes ethical social life. Here is where there is room for fair-market trading to replace deceit and force in bargaining for goods and services; for hospital ethics committees to emerge as responses to tragic problems hitherto baldly faced in isolation by doctors, by nurses, and by parents; for family counseling and negotiation to precede legal disputes in a family divorce court; and so on. In short, although intelligent creative construction is a primary resource for improving ethics, it does not appear ex nihilo but must be nurtured by the kinds of institutional arrangements we have been discussing. And there are two further positives that synergistically come with such nurturing. The first is improvement in rationality; indeed, according to the account we have developed, improvement in ethics comes about through the improvement of resources for reason, especially the increase in creative constructions, such as institutional ethics committees, and richer critical assessment processes. The second is the broad array of societal benefits that flow from a successful culture of creative, constructive problem solving. The conception of knowledge-based, predominantly competitive economics that dominates our current economic phase inhibits those benefits. Creative construction, in contrast, can develop them—the managerial and policy skills required in continual evolution or revolution, for example, and the intelligent and ethical problem-solving approaches needed to manage the cultural acceptance of change—and extend them to social and political processes at large, in turn enlarging the future possibilities for further rational and ethical improvement. This process forms the natural synergy underlying adaptive resilience.3 8.2  Game Theory Rapoport begins his chapter on critiques of game theory by pointing out that game theory was conceived and almost exclusively developed by mathematicians. So the nature of game theory is not surprising: “The theory can be viewed as a self-contained branch of pure mathematics—a system of theorems built up from a set of postulates … when the validity of the theorems, solutions and so on has been checked, nothing remains to be said in criticism, except possibly about the intrinsic depth or elegance of the theory” (Rapoport 1960: 226). The alluring comfort of the necessary, certain correctness that formal reason produces is understandable. What can nonformal reason say by way

202 

Chapter 8

of demurring? Nothing about the mathematics of a game, but something about the design and redesign of the games we strive to play as the expression of our desired society. The importance of getting the nonformal setting for effective problem solving right is wincingly brought home in Rapoport’s subtle but powerful analysis of a game about trust and jealousy based on Shakespeare’s Othello.4 And the classic formal game of Prisoner’s Dilemma demonstrates the need to restructure the problem to escape the skillfully contrived conundrum. Those two examples are considered in turn. Innocent Desdemona represents her situation in a game in which Othello is assumed to share her moral values and has as his dominant strategy believing her to be innocent. She concludes that her best response is to affirm her innocence. This stance is morally better than falsely admitting guilt in the hope of a pardon, a response that also is of inferior expected utility so long as the probability of Othello’s belief in her innocence is greater than 0.5. But Othello is suspicious of her, whether based on the vice of paranoia or the virtue-turned-vice of respecting the advice of a fellow officer whose character and motives have not been insightfully appraised. This insures that his representation of their situation is very different from hers, one in which she and he do not share the same moral values. This discordance is expressed in a game where Desdemona has a dominant deception strategy to which his tragic best response is to believe her guilty. Nonformal judgments of similarity and dissimilarity between players can play crucial roles in negotiations and other problem-solving processes. In Othello’s case, differences in nonformal judgments between Desdemona and him lead not just to differences of detail or even just to differently available strategies, but to divergent forms of the problem-solving process in which they take themselves to be participating. Whether to believe Desdemona as a matter of strategic decision making is quite different from, though subtly related to, whether to believe her as part of conceptualizing the problem itself. Here we discover a recurring theme: rationality cannot consist simply in the formal playing out of formally structured interactions, as in bargaining games and valid arguments, but fundamentally has more to do with the management of nonformal judgments about the problem-presenting situation, about oneself and others (and the relevant similarities and differences involved), and about the kinds of social interaction involved. The classic formal game known as the Prisoner’s Dilemma displays the same theme. This game and related games create situations in which individual players have conflicting incentives to compete (become individually dominant) and to cooperate (attain collective optimality), with the standard maximization solution disappointingly being the competitive one

Designing Practices, Institutions, and Processes 

203

in which both players lose. For these games the truly rational response is not merely to formally analyze the game and accept the individual maximization solution, but to nonformally devise institutional or cultural circumstances that enable the game to be transformed into one where the cooperative solution becomes available at minimal collective cost. Understanding this response requires more detail about the game. A formal game is an abstract choice situation in which each player makes a choice among available strategies and, in consequence, outcomes with positive or negative payoffs are had. In the story that gives Prisoner’s Dilemma its name, two suspected criminals, P1 and P2, are arrested and charged with grand theft and put incommunicado in separate cells. The police offer to reward a confession with a jail sentence shorter than the maximum sentence of 10 years. If, in the face of this, the prisoners cooperate with one another and remain mutually silent, both get a sentence of 5 years on a lesser charge. If both independently deal with the police by confessing, each gets a reduced sentence of 8 years. But if only one prisoner deals with the police and confesses, that prisoner gets the minimal sentence of 3 years, while the prisoner who does not confess gets the maximum sentence of 10 years. Each prisoner’s options are (a) deal with the police (Dpol) and confess or (b) cooperate with the other prisoner (Cpris) and do not confess. Since each chooses between these options independently (prisoners are kept incommunicado), there are four possible outcomes: Dpol/Dpol, Dpol/Cpris, Cpris/Dpol, Cpris/Cpris. The table below displays the abstract form of the game. The potential outcomes for each of the prisoners are in terms of years in jail, which will be represented by negative numbers to make it clear that in this case the aim of each player is to reduce their number toward zero. Then we have: w = –10 (penalty for unilateral cooperation with the other prisoner by not confessing while that cooperation is betrayed by the other prisoner who deals and confesses), z  =  –8 (reward, from 10 down to 8, for mutual prisoner noncooperation when both prisoners confess), x  =  –5 (reward for mutual prisoner cooperation when both prisoners do not confess), y = –3 (reward for unilateral noncooperation to the one prisoner who confesses and betrays the other, cooperative prisoner). In these terms, and listing outcomes in the order Prisoner 1/ Prisoner 2, the four game outcomes above become, respectively: z/z (Dpol/Dpol), w/y (Dpol/Cpris), y/w (Cpris/Dpol), x/x (Cpris/Cpris). The gist of the dilemma is whether a prisoner should pursue self-interest at the risk of betraying the cooperation of the other prisoner. The condition that creates the dilemma is that w